chore(release): 1.1.0

- playout: run-scoped R2 manifest URIs (candidate a rewrite via rewriteManifestToRun)
- tests: playlist_rewrite_test.go (rewrite + publish integration + slate guard)
- docs: STREAMING, cloudflare-workers-setup, architecture (incl. Level-2 diagram)
- release housekeeping: TODO->DONE, CHANGELOG 1.1.0, version bump

.env.example

@@ -8,7 +8,8 @@ R2_SECRET_ACCESS_KEY=your-r2-secret-key
 R2_BUCKET_NAME=linearvoid-media
 R2_ENDPOINT=https://your-account-id.r2.cloudflarestorage.com
 R2_REGION=auto
-R2_SEGMENT_RETENTION_HOURS=24
+R2_KEEP_RUNS=2
+R2_MAX_RETENTION_BYTES=10737418240
 
 # Cloudflare (for custom domain, cache rules, Workers if used)
 CLOUDFLARE_API_TOKEN=your-cloudflare-api-token
@@ -48,9 +49,10 @@ DATABASE_URL=postgres://linearvoid:linearvoid@db:5432/linearvoid?sslmode=disable
 REDIS_URL=redis://redis:6379/0
 
 # Service Ports (for local docker-compose)
-PLAYOUT_PORT=8080
+# Coolify env: 80/443 (Traefik proxy), 8000 (dashboard/Portainer), 8080 taken.
+PLAYOUT_PORT=8081
 CHAT_PORT=3000
-NGINX_PORT=80
+NGINX_PORT=8082
 
 # Playout configuration
 PLAYLIST_PATH=/app/playlist/playlist.txt

.gitea/workflows/cd.yml

@@ -39,8 +39,8 @@ jobs:
 
           # Playout
           PLAYLIST_PATH=/app/playlist/playlist.txt
-          HLS_SEGMENT_DURATION=6
-          HLS_WINDOW_SIZE=24
+          HLS_SEGMENT_DURATION=2
+          HLS_WINDOW_SIZE=12
           LOG_LEVEL=info
 
           # Cloudflare R2 (required for HLS segment sync)
@@ -69,7 +69,8 @@ jobs:
           ADMIN_SESSION_TTL_HOURS=12
 
           # Nginx
-          NGINX_PORT=80
+          # Host port 80/443 (Coolify Traefik) and 8080 are taken; nginx publishes on 8082.
+          NGINX_PORT=8082
           EOF
 
       - name: Seed Prometheus config on remote host
@@ -245,7 +246,7 @@ jobs:
           DOCKER_HOST: tcp://192.168.88.40:2375
         run: |
           # Wait for services to stabilize
-          sleep 20
+          sleep 30
 
           # Check chat health (via docker exec — IP-agnostic, works on any DOCKER_HOST)
           echo "Checking chat health..."
@@ -270,8 +271,8 @@ jobs:
               echo "Playout is healthy"
               break
             fi
-            echo "Playout not ready yet (attempt $i/6), waiting 5s..."
-            sleep 5
+            echo "Playout not ready yet (attempt $i/6), waiting 20s..."
+            sleep 20
           done
           if ! docker exec linearvoid-playout curl -sf http://localhost:8081/health; then
             echo "Playout failed to become healthy after 6 attempts"

.gitea/workflows/pages-deploy.yml

@@ -1,57 +0,0 @@
-name: Deploy to Cloudflare Pages
-
-on:
-  workflow_run:
-    workflows: [CI]
-    types: [completed]
-  workflow_dispatch: {}
-
-jobs:
-  deploy:
-    if: ${{ github.event_name == 'workflow_dispatch' || github.event.workflow_run.conclusion == 'success' }}
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-
-      - name: Setup Node
-        uses: actions/setup-node@v4
-        with:
-          node-version: 22
-
-      - name: Install dependencies
-        run: cd frontend && npm ci
-
-      - name: Build
-        run: cd frontend && npm run build
-
-      - name: Generate deployment manifest
-        run: |
-          cd frontend/dist
-          form_args="-F manifest={"
-          for file in $(find . -type f ! -name manifest.json | sed 's|^\./||'); do
-            hash=$(sha256sum "$file" | awk '{print $1}')
-            form_args="$form_args\"$file\":\"$hash\","
-          done
-          form_args="${form_args%,}}" # remove trailing comma and close JSON
-          echo "Form arguments: $form_args"
-          echo "FORM_ARGS=$form_args" >> $GITHUB_ENV
-
-      - name: Deploy to Cloudflare Pages
-        env:
-          CLOUDFLARE_API_TOKEN: ${{ secrets.CLOUDFLARE_API_TOKEN }}
-          CLOUDFLARE_ACCOUNT_ID: ${{ secrets.CLOUDFLARE_ACCOUNT_ID }}
-        run: |
-          cd frontend/dist
-          form_args="$FORM_ARGS" # use form arguments generated in previous step
-          # Get short commit hash for deployment metadata
-          COMMIT_HASH=$(git rev-parse --short HEAD 2>/dev/null || echo "unknown")
-          COMMIT_MSG=$(git log -1 --pretty=%B 2>/dev/null | head -1 || echo "Gitea Actions deploy")
-
-          curl -sS -X POST \
-            "https://api.cloudflare.com/client/v4/accounts/${CLOUDFLARE_ACCOUNT_ID}/pages/projects/linearvoid/deployments" \
-            -H "Authorization: Bearer ${CLOUDFLARE_API_TOKEN}" \
-            -F "commit_hash=${COMMIT_HASH}" \
-            -F "commit_message=${COMMIT_MSG}" \
-            -F "commit_dirty=false" \
-            $form_args

.gitea/workflows/workers-deploy.yml

@@ -0,0 +1,41 @@
+name: Deploy to Cloudflare Workers
+
+on:
+  workflow_run:
+    workflows: [CI]
+    types: [completed]
+  workflow_dispatch: {}
+
+jobs:
+  deploy:
+    if: ${{ github.event_name == 'workflow_dispatch' || github.event.workflow_run.conclusion == 'success' }}
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Setup Node
+        uses: actions/setup-node@v4
+        with:
+          node-version: 22
+
+      - name: Install dependencies
+        run: cd frontend && npm ci
+
+      - name: Build
+        env:
+          VITE_PUBLIC_APP_URL: ${{ secrets.VITE_PUBLIC_APP_URL }}
+          VITE_PUBLIC_WS_URL: ${{ secrets.VITE_PUBLIC_WS_URL }}
+          VITE_PUBLIC_HLS_URL: ${{ secrets.VITE_PUBLIC_HLS_URL }}
+          VITE_PUBLIC_ADMIN_DEFAULT_MODERATOR: ${{ secrets.VITE_PUBLIC_ADMIN_DEFAULT_MODERATOR }}
+          VITE_PUBLIC_ADMIN_DEFAULT_ROOM: ${{ secrets.VITE_PUBLIC_ADMIN_DEFAULT_ROOM }}
+          VITE_PUBLIC_ADMIN_MODERATION_LOG_LIMIT: "50"
+        run: cd frontend && npm run build
+
+      - name: Deploy to Cloudflare Workers
+        env:
+          CLOUDFLARE_API_TOKEN: ${{ secrets.CLOUDFLARE_API_TOKEN }}
+          CLOUDFLARE_ACCOUNT_ID: ${{ secrets.CLOUDFLARE_ACCOUNT_ID }}
+        run: |
+          cd frontend
+          npx wrangler deploy

.gitea/workflows/workers-media-deploy.yml

@@ -0,0 +1,35 @@
+name: Deploy Media Worker to Cloudflare
+
+on:
+  workflow_run:
+    workflows: [CI]
+    types: [completed]
+  workflow_dispatch: {}
+
+jobs:
+  deploy:
+    # Production deploy: only on pushes to main (via CI success) or manual main runs.
+    if: ${{ (github.event_name == 'workflow_dispatch' && github.ref == 'refs/heads/main') || (github.event_name == 'workflow_run' && github.event.workflow_run.conclusion == 'success' && github.event.workflow_run.head_branch == 'main') }}
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Setup Node
+        uses: actions/setup-node@v4
+        with:
+          node-version: 22
+
+      - name: Install dependencies
+        run: cd infrastructure/cloudflare && npm ci
+
+      - name: Test worker
+        run: cd infrastructure/cloudflare && npm test
+
+      - name: Deploy to Cloudflare Workers (production)
+        env:
+          CLOUDFLARE_API_TOKEN: ${{ secrets.CLOUDFLARE_API_TOKEN }}
+          CLOUDFLARE_ACCOUNT_ID: ${{ secrets.CLOUDFLARE_ACCOUNT_ID }}
+        run: |
+          cd infrastructure/cloudflare
+          npx wrangler deploy --env production

.github/MEMORY.md

@@ -109,7 +109,7 @@
   - Total: 45 frontend tests passing (2 test files).
 - Added coverage reporting: `vitest.config.ts` with v8 provider, thresholds at 20/15/15/20 (statements/branches/functions/lines).
 - Added `test:coverage` script to `package.json`.
-- Fixed CI: bumped Node from 20 to 22 in `ci.yml` and `pages-deploy.yml` (jsdom@30 requires Node 22+). Added `engines` to `package.json`.
+- Fixed CI: bumped Node from 20 to 22 in `ci.yml` and `workers-deploy.yml` (jsdom@30 requires Node 22+). Added `engines` to `package.json`.
 - Updated AGENTS.md: added Required tracking files table with 13 docs and update cadence.
 - All Go vet, Go tests, frontend test:coverage passing.
 - CORS on media.linearvoid.tv remains the only streaming pipeline gap (Cloudflare R2 config, not code).
@@ -138,7 +138,7 @@
 ## 2026-08-13
 
 - Completed all remaining enhancement tasks (only manual/infrastructure phases left):
-  - **Playout**: automatic media normalization on upload (`media.NormalizeVideo`, `*.normalized.mp4`, statuses `ready`/`needs_normalization`/`normalizing`/`normalization_failed`, `POST /api/admin/media/{id}/normalize`); R2 segment retention (`cleanupOldSegments`, `R2_SEGMENT_RETENTION_HOURS`); seamless looping (playlist SHA-256 fingerprint + `-start_number`); single `start-playout.sh` entrypoint + `-validate` mode; loop/cleanup metrics.
+  - **Playout**: automatic media normalization on upload (`media.NormalizeVideo`, `*.normalized.mp4`, statuses `ready`/`needs_normalization`/`normalizing`/`normalization_failed`, `POST /api/admin/media/{id}/normalize`); R2 segment retention (`cleanupOldRuns`, keep last `R2_KEEP_RUNS` runs, `R2_MAX_RETENTION_BYTES` cap); seamless looping (playlist SHA-256 fingerprint + `-start_number`); single `start-playout.sh` entrypoint + `-validate` mode; loop/cleanup metrics.
   - **Application configuration**: thread-safe JSON runtime config store (`backend/chat/src/config/runtime.go`) shared with playout via `/app/playlist/config.json`; `GET/PUT /api/admin/config`; admin UI config form.
   - **Matomo analytics**: public `GET /api/matomo` + viewer pageview snippet in `index.html`.
   - **Chat user accounts**: register/login/logout/session endpoints, bcrypt hashing, `linearvoid_user_session` cookie, viewer sign-in UI.

.github/instructions/BACKLOG.md

@@ -2,12 +2,66 @@
 
 > All current work is planned in [TODO](./TODO.md) (2026-08-13). This file lists backlog items; new unplanned work is added here.
 
+## Setup phases remaining
+
+### Phase 5: complete manual Cloudflare and external-service setup
+
+- [ ] Apply Cloudflare Workers, Tiered Cache, WAF, rate limiting, and HTTPS/HSTS settings in the dashboard
+- [ ] Cloudflare Workers: link custom domain and cut over frontend hosting
+- [ ] Enable Smart Tiered Cache; validate `Cf-Cache-Status: HIT` from upper tier
+- [ ] WAF: enable OWASP Managed Ruleset (Simulate → review 48h → Block), enable Bot Fight Mode
+- [ ] Rate limiting: `/api/admin/login` + `/api/admin/*` → 10 req/10s per IP, Challenge action
+- [ ] HTTPS/WSS: Always Use HTTPS, HSTS max-age=31536000 includeSubDomains preload, min TLS 1.2; enforce at edge and origin
+- [ ] UptimeRobot: add HTTP monitor for `https://linearvoid.tv/health` (5min interval)
+- [ ] UptimeRobot: add keyword monitor for `https://linearvoid.tv/hls/playlist.m3u8` (keyword `#EXTM3U`, 5min)
+- [ ] UptimeRobot: alert on 2 consecutive failures via email + Discord/Slack webhook
+- [ ] Capture evidence in `docs/operations/cloudflare-uptimerobot-checklist.md`
+
+### Phase 6: rehearse production deployment end to end
+
+- [ ] Provision Hetzner CX22 (2 vCPU, 4 GB, 40 GB) in Ashburn VA, Ubuntu 24.04
+- [ ] Install Docker Engine + Compose plugin; open firewall: 80/443 public, 2375 runner-only
+- [ ] Configure Docker daemon for remote TLS access (`-H tcp://0.0.0.0:2376` with certs)
+- [ ] Create `/srv/linearvoid/data/{playlist,media,hls,chat,slate}` directories
+- [ ] Add `CI_PROD_*` secrets to Gitea (DOCKER_HOST, R2 keys, admin password, DB password, Grafana password)
+- [ ] Trigger CD workflow with `DOCKER_HOST` pointing to Hetzner IP
+- [ ] Verify all 11 containers healthy: `/health`, `/hls/playlist.m3u8`, WebSocket
+- [ ] Update Cloudflare A record `linearvoid.tv` → Hetzner IP, SSL Full (strict); test HTTPS + WSS
+- [ ] Smoke test: viewer SPA loads, HLS plays, chat connects, admin login → upload → playlist → playout
+- [ ] Set up Prometheus alerting notifications + UptimeRobot on production URL
+- [ ] Run `scripts/backup-db.sh` + `scripts/verify-backups.sh` on production host
+- [ ] Document production IP, SSH access, rollback procedure, scaling thresholds
+
+### Phase 7: baseline scaling and performance
+
+- [ ] Execute WebSocket load tests — ramp 10→100→500→1000 connections; measure p95 latency + errors
+- [ ] Execute admin API benchmarks — `wrk`/`k6` against `/api/admin/*`; target p95 <1s
+- [ ] Measure manifest TTFB and cache-hit distribution from multiple US regions (WebPageTest 3+ regions, <4s warm cache)
+- [ ] Tune manifest TTL to 0.5× segment duration; validate `Cf-Cache-Status` distribution
+- [ ] Track cache-hit ratio and R2 ops in production (Cloudflare GraphQL → Grafana panel, alert <90%)
+- [ ] Capacity triggers: Prometheus rules for cache-miss >10%/15min, CPU >80%/10min
+- [ ] Plan Phase 2 service separation (playout CPU, chat latency, DB state, monitoring)
+- [ ] Plan Phase 3 multi-origin resilience (second Hetzner location, warm standby, DNS failover, DB/Redis replication)
+
+### Phase 8: close out remaining test and validation gaps
+
+- [ ] Admin frontend: user moderation tools tests (mute/ban/report, view moderation logs)
+- [ ] Frontend: coverage for all critical user + admin components and interactions
+- [ ] Frontend: document test cases and expected outcomes
+- [ ] Backend R2: test retrieval of segments/manifests via Cloudflare edge
+- [ ] Backend R2: validate edge cache hit/miss behavior with repeat requests (`scripts/validate-edge-cache.sh`)
+
+### Playout
+
+- [ ] Configure an R2 lifecycle rule on the `hls/` prefix as defense-in-depth (manual)
+- [ ] Full-loop E2E (2-file playlist, assert `Looping...` with no `CRASHED` across the boundary) — manual via `scripts/test-playout-e2e.sh`
+
 ## Cloudflare Configuration
 
-### DNS & Pages
+### DNS & Workers
 
-- [ ] Cloudflare Pages: link custom domain (manual)
-- [ ] Create Cloudflare Pages project and cut over frontend hosting (manual)
+- [ ] Cloudflare Workers: link custom domain (manual)
+- [ ] Create Cloudflare Worker and cut over frontend hosting (manual)
 
 ### Security & TLS
 
@@ -27,28 +81,6 @@
 - [ ] UptimeRobot: add keyword monitor for `https://linearvoid.tv/hls/playlist.m3u8` (keyword: `#EXTM3U`, 5min)
 - [ ] Alert on 2 consecutive failures, notify via email + Discord/Slack webhook
 
-## Frontend Tests
-
-### User Frontend
-
-- [ ] Test coverage for all critical frontend components and interactions
-- [ ] Add code coverage reporting and enforce minimum coverage threshold in CI/CD pipeline
-- [ ] Document test cases and expected outcomes
-
-### Admin Frontend
-
-- [ ] Test coverage for all critical admin UI components and interactions
-- [ ] Add code coverage reporting and enforce minimum coverage threshold in CI/CD pipeline
-- [ ] Document test cases and expected outcomes
-
-## Backend Integration Tests
-
-### Cloudflare R2
-
-- [ ] Test retrieval of segments and manifests from R2 via Cloudflare edge
-- [ ] Validate cache hit/miss behavior and performance metrics
-- [ ] Validate edge cache hit/miss behavior with repeat requests — script `scripts/validate-edge-cache.sh` created
-
 ## External Services
 
 > Evidence tracker: `docs/operations/cloudflare-uptimerobot-checklist.md`
@@ -102,12 +134,3 @@
 - [ ] Capacity triggers: Prometheus rules for cache-miss >10%/15min, CPU >80%/10min
 - [ ] Plan Phase 2 service separation — document boundaries (playout CPU, chat latency, DB state, monitoring), estimate per-service resources
 - [ ] Plan Phase 3 multi-origin resilience — research second Hetzner location, design warm standby + DNS failover + DB/Redis replication
-
-## Feature List (Future Enhancements)
-
-- [ ] Admin backend
-  - [ ] Analytics: implement a settings entry to set matomo url and page id
-- [ ] User frontend
-  - [ ] Chat: add user registration and login flows
-  - [ ] Chat: private DM history and search
-  - [ ] Analytics: add javascript snippet to send pageview events to matomo

.github/instructions/TODO.md

@@ -2,66 +2,38 @@
 
 > Completed tasks moved to [DONE](./DONE.md). Remaining tasks in [BACKLOG](./BACKLOG.md).
 
-## Chat improvements (remaining)
+## Playout / R2 manifest–segment layout mismatch
 
-- [ ] Manual smoke: viewer join/leave shows a 3s flash (not a chat row); reconnect <5s shows no duplicate; admin chat-log table has no system rows
+**Chosen approach: candidate (a)** — rewrite the live manifest so its media URIs point into the current run-scoped folder before the atomic R2 publish. Go change in `backend/playout/src/playlist.go`. The `linearvoid-media` Worker stays a faithful path→R2 mapper (no serve-time rewrite).
 
-## Playout improvements (remaining)
+### Remaining — Validation
 
-- [ ] Test in Firefox: live playback, rebuffer/recovery, chat sync; record quirks in `docs/STREAMING.md`
-- [ ] Rebuild + run playout; confirm `playlist.m3u8` lists ~12 segments at 2s each
-- [ ] Manual: verify viewer latency and cache-hit behavior after the change
+- [ ] Run playout (local or test lab), then:
+  - `curl https://media.linearvoid.tv/hls/playlist.m3u8` → 200, short TTL.
+  - Every URI in the manifest resolves `200` with `Cache-Control: public, max-age=31536000, immutable` (via the `linearvoid-media` Worker).
+  - Confirm manifest refresh + `CF-Cache-Status` progression and no 404 on segment fetch across a run loop.
+- [ ] Verify slate fallback still serves correctly after the rewrite change.
 
-## Setup phases remaining
+> Implementation (candidate a), Go unit tests, and documentation are complete — see [DONE.md](./DONE.md).
 
-### Phase 5: complete manual Cloudflare and external-service setup
+## Frontend Tests
 
-- [ ] Apply Cloudflare Pages, Tiered Cache, WAF, rate limiting, and HTTPS/HSTS settings in the dashboard
-- [ ] Cloudflare Pages: link custom domain and cut over frontend hosting
-- [ ] Enable Smart Tiered Cache; validate `Cf-Cache-Status: HIT` from upper tier
-- [ ] WAF: enable OWASP Managed Ruleset (Simulate → review 48h → Block), enable Bot Fight Mode
-- [ ] Rate limiting: `/api/admin/login` + `/api/admin/*` → 10 req/10s per IP, Challenge action
-- [ ] HTTPS/WSS: Always Use HTTPS, HSTS max-age=31536000 includeSubDomains preload, min TLS 1.2; enforce at edge and origin
-- [ ] UptimeRobot: add HTTP monitor for `https://linearvoid.tv/health` (5min interval)
-- [ ] UptimeRobot: add keyword monitor for `https://linearvoid.tv/hls/playlist.m3u8` (keyword `#EXTM3U`, 5min)
-- [ ] UptimeRobot: alert on 2 consecutive failures via email + Discord/Slack webhook
-- [ ] Capture evidence in `docs/operations/cloudflare-uptimerobot-checklist.md`
+### User Frontend
 
-### Phase 6: rehearse production deployment end to end
+- [ ] Test coverage for all critical frontend components and interactions
+- [ ] Add code coverage reporting and enforce minimum coverage threshold in CI/CD pipeline
+- [ ] Document test cases and expected outcomes
 
-- [ ] Provision Hetzner CX22 (2 vCPU, 4 GB, 40 GB) in Ashburn VA, Ubuntu 24.04
-- [ ] Install Docker Engine + Compose plugin; open firewall: 80/443 public, 2375 runner-only
-- [ ] Configure Docker daemon for remote TLS access (`-H tcp://0.0.0.0:2376` with certs)
-- [ ] Create `/srv/linearvoid/data/{playlist,media,hls,chat,slate}` directories
-- [ ] Add `CI_PROD_*` secrets to Gitea (DOCKER_HOST, R2 keys, admin password, DB password, Grafana password)
-- [ ] Trigger CD workflow with `DOCKER_HOST` pointing to Hetzner IP
-- [ ] Verify all 11 containers healthy: `/health`, `/hls/playlist.m3u8`, WebSocket
-- [ ] Update Cloudflare A record `linearvoid.tv` → Hetzner IP, SSL Full (strict); test HTTPS + WSS
-- [ ] Smoke test: viewer SPA loads, HLS plays, chat connects, admin login → upload → playlist → playout
-- [ ] Set up Prometheus alerting notifications + UptimeRobot on production URL
-- [ ] Run `scripts/backup-db.sh` + `scripts/verify-backups.sh` on production host
-- [ ] Document production IP, SSH access, rollback procedure, scaling thresholds
+### Admin Frontend
 
-### Phase 7: baseline scaling and performance
+- [ ] Test coverage for all critical admin UI components and interactions
+- [ ] Add code coverage reporting and enforce minimum coverage threshold in CI/CD pipeline
+- [ ] Document test cases and expected outcomes
 
-- [ ] Execute WebSocket load tests — ramp 10→100→500→1000 connections; measure p95 latency + errors
-- [ ] Execute admin API benchmarks — `wrk`/`k6` against `/api/admin/*`; target p95 <1s
-- [ ] Measure manifest TTFB and cache-hit distribution from multiple US regions (WebPageTest 3+ regions, <4s warm cache)
-- [ ] Tune manifest TTL to 0.5× segment duration; validate `Cf-Cache-Status` distribution
-- [ ] Track cache-hit ratio and R2 ops in production (Cloudflare GraphQL → Grafana panel, alert <90%)
-- [ ] Capacity triggers: Prometheus rules for cache-miss >10%/15min, CPU >80%/10min
-- [ ] Plan Phase 2 service separation (playout CPU, chat latency, DB state, monitoring)
-- [ ] Plan Phase 3 multi-origin resilience (second Hetzner location, warm standby, DNS failover, DB/Redis replication)
+## Backend Integration Tests
 
-### Phase 8: close out remaining test and validation gaps
+### Cloudflare R2
 
-- [ ] Admin frontend: user moderation tools tests (mute/ban/report, view moderation logs)
-- [ ] Frontend: coverage for all critical user + admin components and interactions
-- [ ] Frontend: document test cases and expected outcomes
-- [ ] Backend R2: test retrieval of segments/manifests via Cloudflare edge
-- [ ] Backend R2: validate edge cache hit/miss behavior with repeat requests (`scripts/validate-edge-cache.sh`)
-
-### Playout
-
-- [ ] Configure an R2 lifecycle rule on the `hls/` prefix as defense-in-depth (manual)
-- [ ] Full-loop E2E (2-file playlist, assert `Looping...` with no `CRASHED` across the boundary) — manual via `scripts/test-playout-e2e.sh`
+- [ ] Test retrieval of segments and manifests from R2 via Cloudflare edge
+- [ ] Validate cache hit/miss behavior and performance metrics
+- [ ] Validate edge cache hit/miss behavior with repeat requests — script `scripts/validate-edge-cache.sh` created

.prettierrc

@@ -0,0 +1,7 @@
+{
+  "tabWidth": 2,
+  "useTabs": false,
+  "printWidth": 120,
+  "endOfLine": "lf",
+  "proseWrap": "preserve"
+}

CHANGELOG.md

@@ -2,24 +2,35 @@
 
 <!-- markdownlint-disable MD024 -->
 
-## [Unreleased]
+## [1.1.0] - 2026-09-03
 
 ### Added
 
-- **Transient presence flash**: Join/leave announcements now appear as a 3-second flash at the top of the chat panel (`#chat-flash`) with a fade-out, instead of permanent chat rows. Presence indicators still update from the same events.
+- **Cloudflare media edge worker (`linearvoid-media`)**: new self-contained project under `infrastructure/cloudflare/` — an R2-bound Worker owning `media.linearvoid.tv` that enforces immutable segment caching (`public, max-age=31536000, immutable`) and short-TTL live `.m3u8`, with HLS CORS/OPTIONS, byte-Range, and a `/health` probe. Includes Vitest unit tests + Miniflare/workerd R2 integration tests and production deploy via `.gitea/workflows/workers-media-deploy.yml`. Replaces the ad-hoc `worker.js`.
+- **Bash script ports**: Linux/macOS equivalents of the Windows ops/load-test scripts (`cache-hit-ratio`, `check-env`, `check-tokens`, loadtest run/summarize); `.ps1` files retained.
+- **Transient presence flash**: join/leave events show as a 3-second `#chat-flash` instead of permanent chat rows.
 
 ### Changed
 
-- **System messages no longer persisted**: The chat backend no longer stores `system`-type messages; chat history and the admin chat-log table only contain real chat messages (moderation actions remain in the audit log).
-- **Reconnect grace period**: Disconnects schedule a delayed "left" announcement (5s); a reconnect within 5s cancels it and skips "joined", eliminating join/leave spam on brief reconnects.
-- **HLS segment duration 2s**: Default `HLS_SEGMENT_DURATION` changed from 6 to 2 and `HLS_WINDOW_SIZE` from 24 to 12 (2s × 12 = 24s live window), reducing end-to-end latency; the 2s GOP was already in place.
-- **Chat font sizing**: `.message-text` set to `0.95rem` (slightly smaller than the default, larger than the `0.83rem` system/flash text).
+- **Cloudflare Workers hosting**: frontend moved from Cloudflare Pages to the `linearvoid-frontend` static-assets Worker; docs now describe `linearvoid-frontend` (SPA) and `linearvoid-media` (edge HLS), correcting stale `worker.js` and R2 custom-domain wording.
+- **Docs de-Windows'd & condensed**: bash-first guides for `SETUP`, `DEPLOYMENT`, `API`, `MONITORING`, `STREAMING`, `COOLIFY`, incident runbook, scaling playbook, architecture README, and Prometheus metrics inventory.
+- **Compose port de-collision**: playout on host port `8081`, nginx moved to `8082` (`NGINX_PORT` default updated) to avoid Coolify-reserved ports.
+- **Playout**: R2 retention now by run count + byte cap (`R2_KEEP_RUNS` default 2, `R2_MAX_RETENTION_BYTES` default 10 GiB, new gauges) replacing the 24h time-based cleanup; default `HLS_SEGMENT_DURATION` 2s / `HLS_WINDOW_SIZE` 12; deduplicated, concurrent startup validation.
+- **Chat/UX**: 5s reconnect grace window (no join/leave spam); `system` messages no longer persisted; smaller `.message-text`; sidebar cleanup (removed DM search + Users toggle, login/register now a modal).
+- **Auth**: viewer session cookie uses `SameSite=None; Secure` over HTTPS (cross-site SPA) and `Lax` over plain HTTP.
+
+### Fixed
+
+- **Run-scoped R2 manifest URIs (playout)**: `publishManifest` now rewrites bare relative media URIs to the current run folder (`r-{runID}/…`) via `rewriteManifestToRun` before the atomic `.tmp` → `CopyObject` publish, so the live `hls/playlist.m3u8` references resolve against the run-scoped segment keys on `media.linearvoid.tv`. Skipped in slate mode. (Live validation pending — see `TODO.md`.)
+- **Cross-origin auth**: login/register API calls now prefix `VITE_PUBLIC_APP_URL` so the Workers-served SPA reaches the app origin.
+- **Duplicate chat history**: `connectWebSocket` clears messages on each (re)connect and the initial connect waits for session resolution.
 
 ### Tests
 
-- WebSocket: `TestHub_Broadcast_SystemNotPersisted`, `TestHub_ReconnectGracePeriod`.
-- Handlers: `TestHandleAdminChatLog_ExcludesSystem`.
-- Frontend: `presence flash` test.
+- WebSocket (`TestHub_Broadcast_SystemNotPersisted`, `TestHub_ReconnectGracePeriod`), handlers (`TestHandleAdminChatLog_ExcludesSystem`), frontend presence flash.
+- Playout: `TestCleanupOldRuns(_SizeCap)`, `TestR2KeepRuns`, `TestR2MaxRetentionBytes`, `TestR2RunFromKey`, `TestValidateWorkersConfig`, `TestValidatePlaylist_Dedup`.
+- Playout (run-scoped manifest rewrite): `TestRewriteManifestToRun`, `TestRewriteManifestToRun_AlreadyPrefixedUnchanged`, `TestRewriteManifestToRun_HandlesTSAndComments`, `TestRewriteManifestToRun_EmptyRunID`, `TestIsMediaURI`, `TestPublishManifest_RewritesToRunScopedURIs`, `TestPublishManifest_SlateNotRewritten`.
+- Cloudflare media worker: Vitest unit (`cacheControlForPath`, `contentTypeForPath`, `parseByteRange`) + Miniflare R2 integration (immutable segment, short-TTL manifest, Range `206`, `404`, OPTIONS, method guard).
 
 ## [1.0.13] - 2026-08-13
 

README.md

@@ -7,7 +7,7 @@
 [![Build Status](https://git.allucanget.biz/zwitschi/linearvoid/actions/workflows/ci.yml/badge.svg)](https://git.allucanget.biz/zwitschi/linearvoid/actions)
 [![Gitea](https://img.shields.io/badge/Gitea-1.27.1-blue)](https://git.allucanget.biz)
 [![License](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE)
-[![Version](https://img.shields.io/badge/version-1.0.13-blue)](https://git.allucanget.biz/zwitschi/linearvoid/releases)
+[![Version](https://img.shields.io/badge/version-1.1.0-blue)](https://git.allucanget.biz/zwitschi/linearvoid/releases)
 [![Release Notes](https://img.shields.io/badge/release-notes-green)](CHANGELOG.md)
 
 ## Architecture
@@ -187,7 +187,8 @@ frontend/
 Everything needed to deploy and configure the stack (Docker, Nginx, Cloudflare rules, future Terraform).
 
 - Cloudflare
-  - [worker.js](infrastructure/cloudflare/worker.js) — adds immutable cache headers to HLS segments
+  - [linearvoid-media worker](infrastructure/cloudflare/) — R2-bound edge Worker serving HLS on `media.linearvoid.tv` (immutable segment caching, CORS, Range)
+  - [linearvoid-frontend worker](frontend/wrangler.toml) — serves the static SPA on `linearvoid.tv`
 - Nginx
   - [Dockerfile](infrastructure/nginx/Dockerfile) — builds Nginx with custom config
   - [nginx.conf](infrastructure/nginx/nginx.conf) — reverse proxy, static SPA, HLS delivery
@@ -213,11 +214,13 @@ One-off and CI-friendly scripts for video preparation, R2 sync, backups, and pla
 
 - loadtest/
   - [admin-api.js](scripts/loadtest/admin-api.js) — admin API load test script
-  - [run-loadtests.ps1](scripts/loadtest/run-loadtests.ps1) — PowerShell script to run load tests
-  - [summarize-artillery.ps1](scripts/loadtest/summarize-artillery.ps1) — PowerShell script to summarize Artillery load test results
-  - [summarize-k6.ps1](scripts/loadtest/summarize-k6.ps1) — PowerShell script to summarize k6 load test results
+  - [run-loadtests.sh](scripts/loadtest/run-loadtests.sh) — bash script to run ws and/or admin load tests
+  - [summarize-artillery.sh](scripts/loadtest/summarize-artillery.sh) — summarize Artillery JSON results
+  - [summarize-k6.sh](scripts/loadtest/summarize-k6.sh) — summarize k6 JSON results
   - [ws-chat.yml](scripts/loadtest/ws-chat.yml) — WebSocket chat load test configuration
 
+  PowerShell equivalents (`*.ps1`) of the three load-test scripts are kept alongside them.
+
 ## Getting Started
 
 > **Note:** The `docker compose up` below targets local development. The test lab (192.168.88.40) receives automated CD deploys. Production Hetzner VPS is not yet provisioned — see `DEPLOYMENT.md`.
@@ -262,8 +265,8 @@ See the [architecture docs](docs/architecture/README.md) for detailed architectu
 | `VITE_PUBLIC_HLS_URL`  | `https://media.linearvoid.tv/hls/playlist.m3u8` | HLS playlist URL                 |
 | `ADMIN_USERNAME`       | `admin`                                         | Admin panel username             |
 | `ADMIN_PASSWORD`       | —                                               | Admin panel password             |
-| `HLS_SEGMENT_DURATION` | `6`                                             | HLS segment duration in seconds  |
-| `HLS_WINDOW_SIZE`      | `24`                                            | HLS playlist window size         |
+| `HLS_SEGMENT_DURATION` | `2`                                             | HLS segment duration in seconds  |
+| `HLS_WINDOW_SIZE`      | `12`                                            | HLS playlist window size         |
 
 Full reference: [`.env.example`](.env.example).
 

backend/chat/src/auth/auth.go

@@ -88,29 +88,48 @@ func CookieSecure(r *http.Request) bool {
 	return r.TLS != nil
 }
 
-// SetCookie writes the admin session cookie.
-func SetCookie(w http.ResponseWriter, r *http.Request, cookieName, token string, expiresAt time.Time) {
+// SameSiteForUserCookie returns the SameSite policy for the viewer session cookie.
+// Over HTTPS it returns SameSite=None so the cookie is sent cross-site from the
+// Workers-served SPA; over plain HTTP (local dev) it falls back to Lax.
+func SameSiteForUserCookie(r *http.Request) http.SameSite {
+	if CookieSecure(r) {
+		return http.SameSiteNoneMode
+	}
+	return http.SameSiteLaxMode
+}
+
+// SetCookie writes a session cookie with the given SameSite policy.
+// SameSite=None is only valid together with Secure, which is enforced here.
+func SetCookie(w http.ResponseWriter, r *http.Request, cookieName, token string, expiresAt time.Time, sameSite http.SameSite) {
+	secure := CookieSecure(r)
+	if sameSite == http.SameSiteNoneMode {
+		secure = true
+	}
 	http.SetCookie(w, &http.Cookie{
 		Name:     cookieName,
 		Value:    token,
 		Path:     "/",
 		HttpOnly: true,
-		SameSite: http.SameSiteStrictMode,
-		Secure:   CookieSecure(r),
+		SameSite: sameSite,
+		Secure:   secure,
 		Expires:  expiresAt,
 		MaxAge:   int(time.Until(expiresAt).Seconds()),
 	})
 }
 
-// ClearCookie removes the admin session cookie.
-func ClearCookie(w http.ResponseWriter, r *http.Request, cookieName string) {
+// ClearCookie removes a session cookie with the given SameSite policy.
+func ClearCookie(w http.ResponseWriter, r *http.Request, cookieName string, sameSite http.SameSite) {
+	secure := CookieSecure(r)
+	if sameSite == http.SameSiteNoneMode {
+		secure = true
+	}
 	http.SetCookie(w, &http.Cookie{
 		Name:     cookieName,
 		Value:    "",
 		Path:     "/",
 		HttpOnly: true,
-		SameSite: http.SameSiteStrictMode,
-		Secure:   CookieSecure(r),
+		SameSite: sameSite,
+		Secure:   secure,
 		Expires:  time.Unix(0, 0),
 		MaxAge:   -1,
 	})

backend/chat/src/auth/auth_test.go

@@ -134,7 +134,7 @@ func TestSetCookie(t *testing.T) {
 	w := httptest.NewRecorder()
 
 	expiresAt := time.Now().Add(1 * time.Hour)
-	SetCookie(w, req, "test_cookie", "abc123", expiresAt)
+	SetCookie(w, req, "test_cookie", "abc123", expiresAt, http.SameSiteStrictMode)
 
 	cookies := w.Result().Cookies()
 	if len(cookies) != 1 {
@@ -165,7 +165,7 @@ func TestClearCookie(t *testing.T) {
 	req := httptest.NewRequest("GET", "/", nil)
 	w := httptest.NewRecorder()
 
-	ClearCookie(w, req, "test_cookie")
+	ClearCookie(w, req, "test_cookie", http.SameSiteStrictMode)
 
 	cookies := w.Result().Cookies()
 	if len(cookies) != 1 {
@@ -183,6 +183,38 @@ func TestClearCookie(t *testing.T) {
 	}
 }
 
+func TestSameSiteForUserCookie(t *testing.T) {
+	req := httptest.NewRequest("GET", "/", nil)
+	req.Header.Set("X-Forwarded-Proto", "https")
+	if got := SameSiteForUserCookie(req); got != http.SameSiteNoneMode {
+		t.Errorf("SameSiteForUserCookie(https) = %v, want None", got)
+	}
+
+	req2 := httptest.NewRequest("GET", "/", nil)
+	if got := SameSiteForUserCookie(req2); got != http.SameSiteLaxMode {
+		t.Errorf("SameSiteForUserCookie(http) = %v, want Lax", got)
+	}
+}
+
+func TestSetCookie_NoneForcesSecure(t *testing.T) {
+	req := httptest.NewRequest("GET", "/", nil)
+	w := httptest.NewRecorder()
+	expiresAt := time.Now().Add(1 * time.Hour)
+	SetCookie(w, req, "test_cookie", "abc123", expiresAt, http.SameSiteNoneMode)
+
+	cookies := w.Result().Cookies()
+	if len(cookies) != 1 {
+		t.Fatalf("expected 1 cookie, got %d", len(cookies))
+	}
+	c := cookies[0]
+	if c.SameSite != http.SameSiteNoneMode {
+		t.Errorf("cookie SameSite = %v, want None", c.SameSite)
+	}
+	if !c.Secure {
+		t.Error("cookie Secure = false, want true (SameSite=None requires Secure)")
+	}
+}
+
 func TestGetTokenFromCookie(t *testing.T) {
 	req := httptest.NewRequest("GET", "/", nil)
 	req.AddCookie(&http.Cookie{Name: "session", Value: "token123"})

backend/chat/src/handlers/handlers.go

@@ -291,7 +291,7 @@ func HandleAdminLogin(deps *Dependencies) http.HandlerFunc {
 
 		expiresAt := time.Now().Add(config.AdminSessionTTL())
 		deps.SessionStore.Add(token, expiresAt)
-		auth.SetCookie(w, r, config.AdminSessionCookieName, token, expiresAt)
+		auth.SetCookie(w, r, config.AdminSessionCookieName, token, expiresAt, http.SameSiteStrictMode)
 
 		w.Header().Set("Content-Type", "application/json")
 		json.NewEncoder(w).Encode(AdminSessionResponse{Authenticated: true, Username: config.AdminUsername()})
@@ -309,7 +309,7 @@ func HandleAdminLogout(deps *Dependencies) http.HandlerFunc {
 		if token, ok := auth.GetTokenFromCookie(r, config.AdminSessionCookieName); ok {
 			deps.SessionStore.Remove(token)
 		}
-		auth.ClearCookie(w, r, config.AdminSessionCookieName)
+		auth.ClearCookie(w, r, config.AdminSessionCookieName, http.SameSiteStrictMode)
 
 		w.Header().Set("Content-Type", "application/json")
 		json.NewEncoder(w).Encode(AdminSessionResponse{Authenticated: false})
@@ -368,7 +368,7 @@ func startUserSession(deps *Dependencies, w http.ResponseWriter, r *http.Request
 	}
 	expiresAt := time.Now().Add(config.UserSessionTTL())
 	deps.UserSessions.Add(token, username, expiresAt)
-	auth.SetCookie(w, r, userSessionCookieName, token, expiresAt)
+	auth.SetCookie(w, r, userSessionCookieName, token, expiresAt, auth.SameSiteForUserCookie(r))
 }
 
 // HandleUserRegister creates a new chat user and signs them in.
@@ -474,7 +474,7 @@ func HandleUserLogout(deps *Dependencies) http.HandlerFunc {
 		if token, ok := auth.GetTokenFromCookie(r, userSessionCookieName); ok {
 			deps.UserSessions.Remove(token)
 		}
-		auth.ClearCookie(w, r, userSessionCookieName)
+		auth.ClearCookie(w, r, userSessionCookieName, auth.SameSiteForUserCookie(r))
 		w.Header().Set("Content-Type", "application/json")
 		json.NewEncoder(w).Encode(map[string]bool{"ok": true})
 	}

backend/playout/src/playlist.go

@@ -11,8 +11,11 @@ import (
 	"os"
 	"os/exec"
 	"path/filepath"
+	"runtime"
+	"sort"
 	"strconv"
 	"strings"
+	"sync"
 	"time"
 
 	"github.com/aws/aws-sdk-go-v2/aws"
@@ -41,6 +44,9 @@ var (
 	deleteObjects = func(client *s3.Client, input *s3.DeleteObjectsInput) (*s3.DeleteObjectsOutput, error) {
 		return client.DeleteObjects(context.TODO(), input)
 	}
+	copyObject = func(client *s3.Client, input *s3.CopyObjectInput) (*s3.CopyObjectOutput, error) {
+		return client.CopyObject(context.TODO(), input)
+	}
 	bucket             = os.Getenv("R2_BUCKET_NAME")
 	playoutMode        = "starting" // "live", "slate", "degraded"
 	restartCount       = 0
@@ -66,6 +72,10 @@ var (
 	// R2 retention cleanup counter
 	r2CleanupObjectsTotal int64
 
+	// R2 retention gauges
+	r2RunsRetained  int64
+	r2BytesRetained int64
+
 	// Continuous playout loop metrics
 	loopsTotal      int64
 	lastRunDuration time.Duration
@@ -78,6 +88,10 @@ const (
 	// Cache-Control header values for R2 uploads.
 	segmentCacheControl  = "public, max-age=31536000, immutable"
 	manifestCacheControl = "public, max-age=1, s-maxage=2, must-revalidate"
+
+	// R2 retention policy.
+	defaultR2KeepRuns          = 2
+	defaultR2MaxRetentionBytes = int64(10 * 1024 * 1024 * 1024) // 10 GiB
 )
 
 // envWithDefault returns the value of the environment variable or the default if not set.
@@ -174,6 +188,8 @@ func watchAndPublishSegments(ctx context.Context, hlsOutputDir string) {
 }
 
 // publishManifest uploads the manifest atomically: temp key first, then copy to final key (STREAMING.md contract).
+// Bare media URIs in the manifest are rewritten to the run-scoped R2 layout
+// (r-{runID}/…) so that every referenced segment/init resolves on the media host.
 func publishManifest(localManifestPath string) error {
 	if r2Client == nil {
 		if err := initR2Client(); err != nil {
@@ -181,13 +197,33 @@ func publishManifest(localManifestPath string) error {
 		}
 	}
 
+	data, err := os.ReadFile(localManifestPath)
+	if err != nil {
+		return fmt.Errorf("read manifest failed: %w", err)
+	}
+
+	// Rewrite bare relative media URIs to the current run folder so the manifest
+	// matches the run-scoped segment keys uploaded by watchAndPublishSegments /
+	// syncLatestToR2 (candidate (a)). Slate mode is skipped: its manifest references
+	// slate assets that are not published under the live run prefix.
+	if playoutMode != "slate" {
+		data = rewriteManifestToRun(runID, data)
+	}
+
+	// Stage the (rewritten) manifest locally so the existing SyncToR2 path is reused.
+	staging := localManifestPath + ".r2staging"
+	if err := os.WriteFile(staging, data, 0o644); err != nil {
+		return fmt.Errorf("write staging manifest failed: %w", err)
+	}
+	defer os.Remove(staging)
+
 	// 1. Upload to temp key
-	if err := SyncToR2(localManifestPath, "hls/playlist.m3u8.tmp"); err != nil {
+	if err := SyncToR2(staging, "hls/playlist.m3u8.tmp"); err != nil {
 		return fmt.Errorf("temp manifest upload failed: %w", err)
 	}
 
 	// 2. Atomic copy to final key
-	_, err := r2Client.CopyObject(context.TODO(), &s3.CopyObjectInput{
+	_, err = copyObject(r2Client, &s3.CopyObjectInput{
 		Bucket:     aws.String(bucket),
 		CopySource: aws.String(bucket + "/hls/playlist.m3u8.tmp"),
 		Key:        aws.String("hls/playlist.m3u8"),
@@ -199,6 +235,55 @@ func publishManifest(localManifestPath string) error {
 	return nil
 }
 
+// isMediaURI reports whether s names an HLS media file (a segment or init file).
+func isMediaURI(s string) bool {
+	name := s
+	if i := strings.IndexByte(name, '?'); i >= 0 {
+		name = name[:i]
+	}
+	switch strings.ToLower(filepath.Ext(name)) {
+	case ".m4s", ".ts", ".mp4", ".m4a":
+		return true
+	}
+	return false
+}
+
+// rewriteManifestToRun prefixes bare relative media URIs in an HLS manifest with the
+// run folder, e.g. "seg-000001.m4s" -> "r-20260903T…/seg-000001.m4s" and
+// #EXT-X-MAP:URI="init.mp4" -> URI="r-20260903T…/init.mp4". This keeps the published
+// manifest consistent with the run-scoped R2 layout (STREAMING.md contract). Directives,
+// comments, blank lines, and URIs that already contain a path are left unchanged.
+// Returns data unchanged when runID is empty.
+func rewriteManifestToRun(runID string, data []byte) []byte {
+	if runID == "" {
+		return data
+	}
+	prefix := runID + "/"
+	lines := strings.Split(string(data), "\n")
+	out := make([]string, 0, len(lines))
+	for _, line := range lines {
+		trimmed := strings.TrimSpace(line)
+		switch {
+		case strings.HasPrefix(trimmed, "#EXT-X-MAP:URI="):
+			if open := strings.Index(trimmed, `URI="`); open >= 0 {
+				rest := trimmed[open+len(`URI="`):]
+				if closeIdx := strings.IndexByte(rest, '"'); closeIdx > 0 {
+					uri := rest[:closeIdx]
+					if uri != "" && !strings.Contains(uri, "/") {
+						line = strings.Replace(line, `URI="`+uri+`"`, `URI="`+prefix+uri+`"`, 1)
+					}
+				}
+			}
+			out = append(out, line)
+		case trimmed != "" && !strings.HasPrefix(trimmed, "#") && !strings.Contains(trimmed, "/") && isMediaURI(trimmed):
+			out = append(out, prefix+trimmed)
+		default:
+			out = append(out, line)
+		}
+	}
+	return []byte(strings.Join(out, "\n"))
+}
+
 // LoadPlaylist reads a playlist file and returns a list of VideoFile entries
 func LoadPlaylist(playlistPath string) ([]VideoFile, error) {
 	file, err := os.Open(playlistPath)
@@ -229,16 +314,32 @@ func LoadPlaylist(playlistPath string) ([]VideoFile, error) {
 // ValidateVideo uses ffprobe to enforce codec, fps, and keyframe interval (ADR-001)
 func ValidateVideo(videoPath string) (VideoFile, error) {
 	v := VideoFile{Path: videoPath}
-	var errors []string
 
-	// Run ffprobe
-	cmd := exec.Command("ffprobe", "-v", "quiet", "-print_format", "json", "-show_format", "-show_streams", videoPath)
-	output, err := cmd.Output()
-	if err != nil {
-		v.Validation = "ffprobe failed: " + err.Error()
-		return v, fmt.Errorf("ffprobe: %w", err)
+	// Run the stream/format probe and the keyframe probe concurrently: they are
+	// independent, and the keyframe probe (-show_frames) is the expensive one.
+	var streamOut []byte
+	var streamErr error
+	var keyframeErr error
+	var wg sync.WaitGroup
+	wg.Add(2)
+	go func() {
+		defer wg.Done()
+		cmd := exec.Command("ffprobe", "-v", "quiet", "-print_format", "json", "-show_format", "-show_streams", videoPath)
+		streamOut, streamErr = cmd.Output()
+	}()
+	go func() {
+		defer wg.Done()
+		keyframeErr = validateKeyframeInterval(videoPath)
+	}()
+	wg.Wait()
+
+	if streamErr != nil {
+		v.Validation = "ffprobe failed: " + streamErr.Error()
+		return v, fmt.Errorf("ffprobe: %w", streamErr)
 	}
 
+	var errors []string
+
 	// Parse JSON
 	var probe struct {
 		Streams []struct {
@@ -256,7 +357,7 @@ func ValidateVideo(videoPath string) (VideoFile, error) {
 			Duration string `json:"duration"`
 		} `json:"format"`
 	}
-	if err := json.Unmarshal(output, &probe); err != nil {
+	if err := json.Unmarshal(streamOut, &probe); err != nil {
 		v.Validation = "ffprobe JSON parse failed: " + err.Error()
 		return v, fmt.Errorf("ffprobe json: %w", err)
 	}
@@ -317,8 +418,7 @@ func ValidateVideo(videoPath string) (VideoFile, error) {
 		return v, fmt.Errorf("validation failed: %s", v.Validation)
 	}
 
-	// Validate keyframe interval <=2s (60 frames at 30fps)
-	keyframeErr := validateKeyframeInterval(videoPath)
+	// Keyframe interval check (probe already completed concurrently above).
 	if keyframeErr != nil {
 		errors = append(errors, keyframeErr.Error())
 	}
@@ -412,21 +512,62 @@ func ValidateVideoStrict(videoPath string) map[string]any {
 	return result
 }
 
-// ValidatePlaylist runs normalization validation on all entries
+// validateWorkers returns the number of concurrent validation workers.
+func validateWorkers() int {
+	n := runtime.NumCPU()
+	if v := os.Getenv("PLAYOUT_VALIDATE_WORKERS"); v != "" {
+		if parsed, err := strconv.Atoi(v); err == nil && parsed > 0 {
+			n = parsed
+		}
+	}
+	return n
+}
+
+// ValidatePlaylist runs normalization validation on all entries in parallel,
+// deduplicating repeated paths so each unique file is probed once.
 func ValidatePlaylist(playlistPath string) ([]VideoFile, error) {
 	videos, err := LoadPlaylist(playlistPath)
 	if err != nil {
 		return nil, err
 	}
 
-	for i := range videos {
-		validated, err := ValidateVideo(videos[i].Path)
-		videos[i] = validated
-		if err != nil {
-			fmt.Printf("Warning: %s - %s\n", videos[i].Path, err)
+	// Collect unique paths in order of first appearance.
+	var order []string
+	seen := make(map[string]bool, len(videos))
+	for _, v := range videos {
+		if !seen[v.Path] {
+			seen[v.Path] = true
+			order = append(order, v.Path)
 		}
 	}
 
+	// Validate unique paths concurrently with a bounded worker pool.
+	results := make(map[string]VideoFile, len(order))
+	var mu sync.Mutex
+	var wg sync.WaitGroup
+	sem := make(chan struct{}, validateWorkers())
+	for _, path := range order {
+		wg.Add(1)
+		sem <- struct{}{}
+		go func(p string) {
+			defer wg.Done()
+			defer func() { <-sem }()
+			vf, err := ValidateVideo(p)
+			if err != nil {
+				fmt.Printf("Warning: %s - %s\n", p, err)
+			}
+			mu.Lock()
+			results[p] = vf
+			mu.Unlock()
+		}(path)
+	}
+	wg.Wait()
+
+	// Map results back onto the playlist (preserving order and duplicates).
+	for i := range videos {
+		videos[i] = results[videos[i].Path]
+	}
+
 	return videos, nil
 }
 
@@ -623,11 +764,24 @@ func main() {
 	initRunID()
 	fmt.Println("Run ID:", runID)
 
-	// Validate playlist before starting (ADR-001 normalization enforcement)
-	if _, err := ValidatePlaylist(playlistPath); err != nil {
+	// Validate playlist before starting (ADR-001 normalization enforcement).
+	// Runs once here; start-playout.sh no longer pre-validates separately.
+	videos, err := ValidatePlaylist(playlistPath)
+	if err != nil {
 		fmt.Println("ERROR: Playlist validation failed:", err)
 		os.Exit(1)
 	}
+	invalid := 0
+	for _, v := range videos {
+		if !v.Valid {
+			fmt.Printf("ERROR: invalid media: %s — %s\n", v.Path, v.Validation)
+			invalid++
+		}
+	}
+	if invalid > 0 {
+		fmt.Printf("ERROR: Playlist validation failed: %d invalid entries\n", invalid)
+		os.Exit(1)
+	}
 
 	// Initialize R2 client (skip if not configured — local HLS via nginx still works)
 	if err := initR2Client(); err != nil {
@@ -658,6 +812,12 @@ func main() {
 		fmt.Fprintln(w, "# HELP linearvoid_playout_r2_cleanup_objects_total Total R2 objects deleted by retention cleanup.")
 		fmt.Fprintln(w, "# TYPE linearvoid_playout_r2_cleanup_objects_total counter")
 		fmt.Fprintf(w, "linearvoid_playout_r2_cleanup_objects_total %d\n", r2CleanupObjectsTotal)
+		fmt.Fprintln(w, "# HELP linearvoid_playout_r2_runs_retained Number of R2 run folders retained by cleanup.")
+		fmt.Fprintln(w, "# TYPE linearvoid_playout_r2_runs_retained gauge")
+		fmt.Fprintf(w, "linearvoid_playout_r2_runs_retained %d\n", r2RunsRetained)
+		fmt.Fprintln(w, "# HELP linearvoid_playout_r2_bytes_retained Total bytes retained in R2 run folders.")
+		fmt.Fprintln(w, "# TYPE linearvoid_playout_r2_bytes_retained gauge")
+		fmt.Fprintf(w, "linearvoid_playout_r2_bytes_retained %d\n", r2BytesRetained)
 		fmt.Fprintln(w, "# HELP linearvoid_playout_loops_total Total clean playlist completions (loops).")
 		fmt.Fprintln(w, "# TYPE linearvoid_playout_loops_total counter")
 		fmt.Fprintf(w, "linearvoid_playout_loops_total %d\n", loopsTotal)
@@ -682,6 +842,15 @@ func main() {
 		// New run ID on each restart for unique R2 segment keys (STREAMING.md)
 		initRunID()
 
+		// Remove run folders older than the retention policy (keep the last N runs,
+		// cap total bytes) now that the next run has started. This never touches the
+		// current or previous run folders.
+		if r2Client != nil {
+			if _, err := cleanupOldRuns(context.Background()); err != nil {
+				fmt.Println("R2 cleanup error:", err)
+			}
+		}
+
 		// Reset HLS output only when the playlist changed, we are recovering from a
 		// crash/slate, or this is the first run. Otherwise continue the segment
 		// sequence so the loop is seamless for viewers (no gap on clean loops).
@@ -834,68 +1003,51 @@ func syncLatestToR2(hlsDir string) error {
 	return nil
 }
 
-// r2SegmentRetentionHours returns the R2 segment retention window in hours.
-func r2SegmentRetentionHours() time.Duration {
-	h, err := strconv.Atoi(envWithDefault("R2_SEGMENT_RETENTION_HOURS", "24"))
-	if err != nil || h <= 0 {
-		h = 24
+// r2KeepRuns returns how many of the most recent run folders to keep in R2.
+func r2KeepRuns() int {
+	n, err := strconv.Atoi(envWithDefault("R2_KEEP_RUNS", "2"))
+	if err != nil || n < 1 {
+		return defaultR2KeepRuns
 	}
-	return time.Duration(h) * time.Hour
+	return n
 }
 
-// cleanupOldSegments deletes R2 objects under the hls/ prefix that are older than
-// the retention window, excluding the live manifest and any non-run keys.
-// It returns the number of objects deleted.
-func cleanupOldSegments(ctx context.Context) (int, error) {
-	if r2Client == nil {
-		if err := initR2Client(); err != nil {
-			return 0, err
-		}
+// r2MaxRetentionBytes returns the maximum total bytes to retain across run folders.
+func r2MaxRetentionBytes() int64 {
+	n, err := strconv.ParseInt(envWithDefault("R2_MAX_RETENTION_BYTES", "10737418240"), 10, 64)
+	if err != nil || n <= 0 {
+		return defaultR2MaxRetentionBytes
 	}
+	return n
+}
 
-	cutoff := time.Now().UTC().Add(-r2SegmentRetentionHours())
-
-	var stale []string
-	var token *string
-	for {
-		output, err := listObjects(r2Client, &s3.ListObjectsV2Input{
-			Bucket:            aws.String(bucket),
-			Prefix:            aws.String("hls/"),
-			ContinuationToken: token,
-		})
-		if err != nil {
-			return 0, err
-		}
-		for _, obj := range output.Contents {
-			key := aws.ToString(obj.Key)
-			if key == "hls/playlist.m3u8" || key == "hls/playlist.m3u8.tmp" {
-				continue
-			}
-			// Only consider segment/init objects under a run directory: hls/r-.../...
-			if !strings.HasPrefix(key, "hls/r-") {
-				continue
-			}
-			if obj.LastModified != nil && obj.LastModified.Before(cutoff) {
-				stale = append(stale, key)
-			}
-		}
-		if output.IsTruncated == nil || !*output.IsTruncated {
-			break
-		}
-		token = output.NextContinuationToken
+// r2RunFromKey extracts the run folder name ("r-...") from an R2 object key.
+func r2RunFromKey(key string) (string, bool) {
+	rest := strings.TrimPrefix(key, "hls/")
+	if rest == key || !strings.HasPrefix(rest, "r-") {
+		return "", false
 	}
-
-	if len(stale) == 0 {
-		return 0, nil
+	if idx := strings.IndexByte(rest, '/'); idx > 0 {
+		return rest[:idx], true
 	}
+	return "", false
+}
 
+// runObjects groups the R2 object keys and total size for a single run folder.
+type runObjects struct {
+	keys []string
+	size int64
+}
+
+// deleteR2Keys deletes the given R2 object keys in batches of up to 1000.
+func deleteR2Keys(keys []string) (int, error) {
 	deleted := 0
-	for i := 0; i < len(stale); i += 1000 {
+	for i := 0; i < len(keys); i += 1000 {
 		end := i + 1000
-		if end > len(stale) {
-			end = len(stale)
+		if end > len(keys) {
+			end = len(keys)
 		}
-		batch := stale[i:end]
+		batch := keys[i:end]
 		objects := make([]types.ObjectIdentifier, len(batch))
 		for j, key := range batch {
 			objects[j] = types.ObjectIdentifier{Key: aws.String(key)}
@@ -911,6 +1063,118 @@ func cleanupOldSegments(ctx context.Context) (int, error) {
 		}
 		deleted += len(batch)
 	}
+	return deleted, nil
+}
+
+// cleanupOldRuns lists run-scoped objects under hls/, keeps the most recent
+// R2_KEEP_RUNS runs, and deletes older runs. It also enforces the
+// R2_MAX_RETENTION_BYTES cap by removing oldest retained runs first, always
+// keeping at least the newest run. The live manifest keys are never deleted.
+func cleanupOldRuns(ctx context.Context) (int, error) {
+	if r2Client == nil {
+		if err := initR2Client(); err != nil {
+			return 0, err
+		}
+	}
+
+	runs := map[string]*runObjects{}
+	var token *string
+	for {
+		output, err := listObjects(r2Client, &s3.ListObjectsV2Input{
+			Bucket:            aws.String(bucket),
+			Prefix:            aws.String("hls/"),
+			ContinuationToken: token,
+		})
+		if err != nil {
+			return 0, err
+		}
+		for _, obj := range output.Contents {
+			key := aws.ToString(obj.Key)
+			if key == "hls/playlist.m3u8" || key == "hls/playlist.m3u8.tmp" {
+				continue
+			}
+			run, ok := r2RunFromKey(key)
+			if !ok {
+				continue
+			}
+			r, exists := runs[run]
+			if !exists {
+				r = &runObjects{}
+				runs[run] = r
+			}
+			r.keys = append(r.keys, key)
+			if obj.Size != nil {
+				r.size += *obj.Size
+			}
+		}
+		if output.IsTruncated == nil || !*output.IsTruncated {
+			break
+		}
+		token = output.NextContinuationToken
+	}
+
+	if len(runs) == 0 {
+		r2RunsRetained = 0
+		r2BytesRetained = 0
+		return 0, nil
+	}
+
+	// Run IDs are ISO timestamps, so lexicographic order equals chronological order.
+	sortedRuns := make([]string, 0, len(runs))
+	for run := range runs {
+		sortedRuns = append(sortedRuns, run)
+	}
+	sort.Strings(sortedRuns) // ascending: oldest -> newest
+
+	deleteRun := map[string]bool{}
+	keepRuns := r2KeepRuns()
+
+	// Mark all but the newest keepRuns runs for deletion.
+	if len(sortedRuns) > keepRuns {
+		for _, run := range sortedRuns[:len(sortedRuns)-keepRuns] {
+			deleteRun[run] = true
+		}
+	}
+
+	// Enforce the total size cap on retained runs (oldest retained first), always
+	// keeping the newest run even if it alone exceeds the cap.
+	maxBytes := r2MaxRetentionBytes()
+	retained := make([]string, 0, keepRuns)
+	var retainedBytes int64
+	for _, run := range sortedRuns {
+		if !deleteRun[run] {
+			retained = append(retained, run)
+			retainedBytes += runs[run].size
+		}
+	}
+	for i := 0; retainedBytes > maxBytes && i < len(retained)-1; i++ {
+		run := retained[i]
+		deleteRun[run] = true
+		retainedBytes -= runs[run].size
+	}
+
+	var toDelete []string
+	for run, del := range deleteRun {
+		if del {
+			toDelete = append(toDelete, runs[run].keys...)
+		}
+	}
+
+	deleted, err := deleteR2Keys(toDelete)
+	if err != nil {
+		return deleted, err
+	}
+
+	keptRuns := int64(0)
+	keptBytes := int64(0)
+	for run, r := range runs {
+		if !deleteRun[run] {
+			keptRuns++
+			keptBytes += r.size
+		}
+	}
+	r2RunsRetained = keptRuns
+	r2BytesRetained = keptBytes
 
 	r2CleanupObjectsTotal += int64(deleted)
 	return deleted, nil
@@ -918,13 +1182,13 @@ func cleanupOldSegments(ctx context.Context) (int, error) {
 
 // runR2CleanupLoop runs retention cleanup once at startup, then hourly.
 func runR2CleanupLoop() {
-	if _, err := cleanupOldSegments(context.Background()); err != nil {
+	if _, err := cleanupOldRuns(context.Background()); err != nil {
 		fmt.Println("R2 cleanup error:", err)
 	}
 	ticker := time.NewTicker(time.Hour)
 	defer ticker.Stop()
 	for range ticker.C {
-		if _, err := cleanupOldSegments(context.Background()); err != nil {
+		if _, err := cleanupOldRuns(context.Background()); err != nil {
 			fmt.Println("R2 cleanup error:", err)
 		}
 	}

backend/playout/src/playlist_rewrite_test.go

@@ -0,0 +1,227 @@
+package main
+
+import (
+	"io"
+	"os"
+	"path/filepath"
+	"testing"
+
+	"github.com/aws/aws-sdk-go-v2/aws"
+	"github.com/aws/aws-sdk-go-v2/service/s3"
+)
+
+const sampleManifest = `#EXTM3U
+#EXT-X-VERSION:7
+#EXT-X-TARGETDURATION:2
+#EXT-X-MEDIA-SEQUENCE:0
+#EXT-X-INDEPENDENT-SEGMENTS
+#EXT-X-MAP:URI="init.mp4"
+#EXTINF:2.000000,
+seg-000000.m4s
+#EXTINF:2.000000,
+seg-000001.m4s
+#EXT-X-ENDLIST
+`
+
+func TestRewriteManifestToRun(t *testing.T) {
+	runID := "r-20260903T120000Z"
+	out := string(rewriteManifestToRun(runID, []byte(sampleManifest)))
+
+	wants := []string{
+		`#EXT-X-MAP:URI="r-20260903T120000Z/init.mp4"`,
+		"r-20260903T120000Z/seg-000000.m4s",
+		"r-20260903T120000Z/seg-000001.m4s",
+		"#EXTM3U",
+		"#EXT-X-VERSION:7",
+		"#EXT-X-TARGETDURATION:2",
+		"#EXT-X-MEDIA-SEQUENCE:0",
+		"#EXT-X-INDEPENDENT-SEGMENTS",
+		"#EXT-X-ENDLIST",
+		"#EXTINF:2.000000,",
+	}
+	for _, want := range wants {
+		if !contains(out, want) {
+			t.Errorf("output missing %q\noutput:\n%s", want, out)
+		}
+	}
+
+	// Media URIs must be prefixed (no bare segments left).
+	for _, bare := range []string{"\nseg-000000.m4s\n", "\nseg-000001.m4s\n", `URI="init.mp4"`} {
+		if contains(out, bare) {
+			t.Errorf("output still contains bare %q\noutput:\n%s", bare, out)
+		}
+	}
+}
+
+func TestRewriteManifestToRun_AlreadyPrefixedUnchanged(t *testing.T) {
+	manifest := "#EXTM3U\n#EXT-X-MAP:URI=\"r-old/init.mp4\"\nseg-000000.m4s\nr-old/seg-000001.m4s\n#EXT-X-ENDLIST\n"
+	out := string(rewriteManifestToRun("r-new", []byte(manifest)))
+
+	if !contains(out, `#EXT-X-MAP:URI="r-old/init.mp4"`) {
+		t.Errorf("EXT-X-MAP already-prefixed URI was rewritten: %s", out)
+	}
+	// A URI line that already contains a path is left alone.
+	if !contains(out, "r-old/seg-000001.m4s") {
+		t.Errorf("already-prefixed segment URI was changed:\n%s", out)
+	}
+	// Bare URI still gets the run prefix.
+	if !contains(out, "r-new/seg-000000.m4s") {
+		t.Errorf("bare segment not prefixed:\n%s", out)
+	}
+}
+
+func TestRewriteManifestToRun_HandlesTSAndComments(t *testing.T) {
+	manifest := "#EXTM3U\n# a comment\nvideo-1.ts\ninit.mp4\n#EXT-X-ENDLIST\n"
+	out := string(rewriteManifestToRun("r-x", []byte(manifest)))
+
+	for _, want := range []string{"r-x/video-1.ts", "r-x/init.mp4", "# a comment", "#EXTM3U", "#EXT-X-ENDLIST"} {
+		if !contains(out, want) {
+			t.Errorf("output missing %q\noutput:\n%s", want, out)
+		}
+	}
+}
+
+func TestRewriteManifestToRun_EmptyRunID(t *testing.T) {
+	out := rewriteManifestToRun("", []byte(sampleManifest))
+	if string(out) != sampleManifest {
+		t.Errorf("empty runID should return input unchanged")
+	}
+}
+
+func TestIsMediaURI(t *testing.T) {
+	for _, ok := range []string{"seg-000001.m4s", "init.mp4", "a.ts", "x.m4a"} {
+		if !isMediaURI(ok) {
+			t.Errorf("isMediaURI(%q) = false, want true", ok)
+		}
+	}
+	for _, notOK := range []string{"playlist.m3u8", "readme.txt", "seg-000001.m4s?x=1"} {
+		// Note: query strings are stripped; '?x=1' on m4s should still be media.
+		if notOK == "seg-000001.m4s?x=1" {
+			continue
+		}
+		if isMediaURI(notOK) {
+			t.Errorf("isMediaURI(%q) = true, want false", notOK)
+		}
+	}
+	if !isMediaURI("seg-000001.m4s?token=abc") {
+		t.Errorf("isMediaURI with query should strip query and be true")
+	}
+}
+
+func TestPublishManifest_RewritesToRunScopedURIs(t *testing.T) {
+	prevClient, prevBucket := r2Client, bucket
+	prevRunID, prevMode := runID, playoutMode
+	prevPut, prevCopy := putObject, copyObject
+	defer func() {
+		r2Client, bucket = prevClient, prevBucket
+		runID, playoutMode = prevRunID, prevMode
+		putObject, copyObject = prevPut, prevCopy
+	}()
+
+	tmpDir := t.TempDir()
+	mp := filepath.Join(tmpDir, "playlist.m3u8")
+	if err := os.WriteFile(mp, []byte(sampleManifest), 0o644); err != nil {
+		t.Fatalf("write manifest: %v", err)
+	}
+
+	bucket = "linearvoid-media"
+	runID = "r-20260903T120000Z"
+	playoutMode = "live"
+	r2Client = &s3.Client{}
+
+	var gotKey, gotCopySource, gotCopyKey string
+	var gotBody []byte
+	putObject = func(_ *s3.Client, in *s3.PutObjectInput) (*s3.PutObjectOutput, error) {
+		gotKey = aws.ToString(in.Key)
+		b, err := io.ReadAll(in.Body)
+		if err != nil {
+			return nil, err
+		}
+		gotBody = b
+		return &s3.PutObjectOutput{}, nil
+	}
+	copyObject = func(_ *s3.Client, in *s3.CopyObjectInput) (*s3.CopyObjectOutput, error) {
+		gotCopySource = aws.ToString(in.CopySource)
+		gotCopyKey = aws.ToString(in.Key)
+		return &s3.CopyObjectOutput{}, nil
+	}
+
+	if err := publishManifest(mp); err != nil {
+		t.Fatalf("publishManifest error: %v", err)
+	}
+
+	if gotKey != "hls/playlist.m3u8.tmp" {
+		t.Errorf("temp upload key = %q, want hls/playlist.m3u8.tmp", gotKey)
+	}
+	if gotCopySource != "linearvoid-media/hls/playlist.m3u8.tmp" {
+		t.Errorf("CopySource = %q, want linearvoid-media/hls/playlist.m3u8.tmp", gotCopySource)
+	}
+	if gotCopyKey != "hls/playlist.m3u8" {
+		t.Errorf("final key = %q, want hls/playlist.m3u8", gotCopyKey)
+	}
+
+	body := string(gotBody)
+	for _, want := range []string{
+		"r-20260903T120000Z/seg-000000.m4s",
+		"r-20260903T120000Z/seg-000001.m4s",
+		`#EXT-X-MAP:URI="r-20260903T120000Z/init.mp4"`,
+	} {
+		if !contains(body, want) {
+			t.Errorf("published manifest missing %q:\n%s", want, body)
+		}
+	}
+	for _, bare := range []string{"\nseg-000000.m4s\n", "\nseg-000001.m4s\n"} {
+		if contains(body, bare) {
+			t.Errorf("published manifest still contains bare %q:\n%s", bare, body)
+		}
+	}
+}
+
+func TestPublishManifest_SlateNotRewritten(t *testing.T) {
+	prevClient, prevBucket := r2Client, bucket
+	prevRunID, prevMode := runID, playoutMode
+	prevPut, prevCopy := putObject, copyObject
+	defer func() {
+		r2Client, bucket = prevClient, prevBucket
+		runID, playoutMode = prevRunID, prevMode
+		putObject, copyObject = prevPut, prevCopy
+	}()
+
+	tmpDir := t.TempDir()
+	mp := filepath.Join(tmpDir, "playlist.m3u8")
+	if err := os.WriteFile(mp, []byte(sampleManifest), 0o644); err != nil {
+		t.Fatalf("write manifest: %v", err)
+	}
+
+	bucket = "linearvoid-media"
+	runID = "r-20260903T120000Z"
+	playoutMode = "slate"
+	r2Client = &s3.Client{}
+
+	var gotBody []byte
+	putObject = func(_ *s3.Client, in *s3.PutObjectInput) (*s3.PutObjectOutput, error) {
+		b, _ := io.ReadAll(in.Body)
+		gotBody = b
+		return &s3.PutObjectOutput{}, nil
+	}
+	copyObject = func(_ *s3.Client, _ *s3.CopyObjectInput) (*s3.CopyObjectOutput, error) {
+		return &s3.CopyObjectOutput{}, nil
+	}
+
+	if err := publishManifest(mp); err != nil {
+		t.Fatalf("publishManifest error: %v", err)
+	}
+
+	if string(gotBody) != sampleManifest {
+		t.Errorf("slate manifest should be published unmodified; got:\n%s", string(gotBody))
+	}
+}
+
+func contains(s, sub string) bool {
+	for i := 0; i+len(sub) <= len(s); i++ {
+		if s[i:i+len(sub)] == sub {
+			return true
+		}
+	}
+	return false
+}

backend/playout/src/playlist_test.go

@@ -465,6 +465,54 @@ func TestValidatePlaylist(t *testing.T) {
 	t.Skip("requires ffprobe mock or test media fixtures")
 }
 
+func TestValidateWorkersConfig(t *testing.T) {
+	os.Unsetenv("PLAYOUT_VALIDATE_WORKERS")
+	if got := validateWorkers(); got < 1 {
+		t.Errorf("validateWorkers() = %d, want >= 1", got)
+	}
+
+	os.Setenv("PLAYOUT_VALIDATE_WORKERS", "3")
+	if got := validateWorkers(); got != 3 {
+		t.Errorf("validateWorkers() = %d, want 3", got)
+	}
+
+	os.Setenv("PLAYOUT_VALIDATE_WORKERS", "invalid")
+	if got := validateWorkers(); got < 1 {
+		t.Errorf("validateWorkers() = %d, want fallback >= 1", got)
+	}
+
+	os.Unsetenv("PLAYOUT_VALIDATE_WORKERS")
+}
+
+func TestValidatePlaylist_Dedup(t *testing.T) {
+	ensureFFProbeInPath(t)
+
+	valid := filepath.Join("..", "testdata", "valid_720p30.mp4")
+	playlistDir := t.TempDir()
+	p := filepath.Join(playlistDir, "playlist.txt")
+	// Same file listed three times to exercise dedup + order preservation.
+	content := fmt.Sprintf("file '%s'\nfile '%s'\nfile '%s'\n", valid, valid, valid)
+	if err := os.WriteFile(p, []byte(content), 0o644); err != nil {
+		t.Fatalf("write playlist: %v", err)
+	}
+
+	videos, err := ValidatePlaylist(p)
+	if err != nil {
+		t.Fatalf("ValidatePlaylist error: %v", err)
+	}
+	if len(videos) != 3 {
+		t.Fatalf("len = %d, want 3", len(videos))
+	}
+	for i, v := range videos {
+		if !v.Valid {
+			t.Errorf("entry %d invalid: %s", i, v.Validation)
+		}
+		if v.Path != valid {
+			t.Errorf("entry %d path = %q, want %q", i, v.Path, valid)
+		}
+	}
+}
+
 func TestSyncToR2(t *testing.T) {
 	prevClient := r2Client
 	prevBucket := bucket
@@ -593,54 +641,95 @@ func TestSyncToR2_NoCacheHeadersForUnknownExt(t *testing.T) {
 	}
 }
 
-func TestR2SegmentRetentionHours(t *testing.T) {
-	os.Setenv("R2_SEGMENT_RETENTION_HOURS", "")
-	if got := r2SegmentRetentionHours(); got != 24*time.Hour {
-		t.Errorf("default = %v, want 24h", got)
+func TestR2KeepRuns(t *testing.T) {
+	os.Setenv("R2_KEEP_RUNS", "")
+	if got := r2KeepRuns(); got != 2 {
+		t.Errorf("default = %d, want 2", got)
 	}
 
-	os.Setenv("R2_SEGMENT_RETENTION_HOURS", "48")
-	if got := r2SegmentRetentionHours(); got != 48*time.Hour {
-		t.Errorf("override = %v, want 48h", got)
+	os.Setenv("R2_KEEP_RUNS", "5")
+	if got := r2KeepRuns(); got != 5 {
+		t.Errorf("override = %d, want 5", got)
 	}
 
-	os.Setenv("R2_SEGMENT_RETENTION_HOURS", "invalid")
-	if got := r2SegmentRetentionHours(); got != 24*time.Hour {
-		t.Errorf("invalid = %v, want 24h fallback", got)
+	os.Setenv("R2_KEEP_RUNS", "0")
+	if got := r2KeepRuns(); got != 2 {
+		t.Errorf("invalid = %d, want 2 fallback", got)
 	}
 
-	os.Unsetenv("R2_SEGMENT_RETENTION_HOURS")
+	os.Unsetenv("R2_KEEP_RUNS")
 }
 
-func TestCleanupOldSegments(t *testing.T) {
+func TestR2MaxRetentionBytes(t *testing.T) {
+	os.Setenv("R2_MAX_RETENTION_BYTES", "")
+	if got := r2MaxRetentionBytes(); got != 10*1024*1024*1024 {
+		t.Errorf("default = %d, want 10GiB", got)
+	}
+
+	os.Setenv("R2_MAX_RETENTION_BYTES", "100")
+	if got := r2MaxRetentionBytes(); got != 100 {
+		t.Errorf("override = %d, want 100", got)
+	}
+
+	os.Setenv("R2_MAX_RETENTION_BYTES", "invalid")
+	if got := r2MaxRetentionBytes(); got != 10*1024*1024*1024 {
+		t.Errorf("invalid = %d, want 10GiB fallback", got)
+	}
+
+	os.Unsetenv("R2_MAX_RETENTION_BYTES")
+}
+
+func TestR2RunFromKey(t *testing.T) {
+	cases := []struct {
+		key string
+		run string
+		ok  bool
+	}{
+		{"hls/r-20260815T120000Z/seg-1.m4s", "r-20260815T120000Z", true},
+		{"hls/r-20260815T120000Z/init.mp4", "r-20260815T120000Z", true},
+		{"hls/playlist.m3u8", "", false},
+		{"hls/playlist.m3u8.tmp", "", false},
+		{"other/r-x/seg.m4s", "", false},
+		{"hls/r-noslash", "", false},
+	}
+	for _, c := range cases {
+		run, ok := r2RunFromKey(c.key)
+		if run != c.run || ok != c.ok {
+			t.Errorf("r2RunFromKey(%q) = (%q, %v), want (%q, %v)", c.key, run, ok, c.run, c.ok)
+		}
+	}
+}
+
+func TestCleanupOldRuns(t *testing.T) {
 	prevClient := r2Client
 	prevBucket := bucket
 	prevList := listObjects
 	prevDelete := deleteObjects
 	prevTotal := r2CleanupObjectsTotal
+	prevRuns := r2RunsRetained
+	prevBytes := r2BytesRetained
 	defer func() {
 		r2Client = prevClient
 		bucket = prevBucket
 		listObjects = prevList
 		deleteObjects = prevDelete
 		r2CleanupObjectsTotal = prevTotal
+		r2RunsRetained = prevRuns
+		r2BytesRetained = prevBytes
 	}()
 
 	bucket = "linearvoid-media"
 	r2Client = &s3.Client{}
 
-	now := time.Now().UTC()
-	old := now.Add(-25 * time.Hour)
-	recent := now.Add(-1 * time.Hour)
-
+	size := int64(1024)
 	listObjects = func(_ *s3.Client, input *s3.ListObjectsV2Input) (*s3.ListObjectsV2Output, error) {
 		return &s3.ListObjectsV2Output{
 			Contents: []types.Object{
-				{Key: aws.String("hls/r-old/seg-1.m4s"), LastModified: &old},
-				{Key: aws.String("hls/r-old/init.mp4"), LastModified: &old},
-				{Key: aws.String("hls/r-new/seg-2.m4s"), LastModified: &recent},
-				{Key: aws.String("hls/playlist.m3u8"), LastModified: &old},
-				{Key: aws.String("hls/playlist.m3u8.tmp"), LastModified: &old},
+				{Key: aws.String("hls/r-20260813T000000Z/seg-1.m4s"), Size: &size},
+				{Key: aws.String("hls/r-20260814T000000Z/seg-2.m4s"), Size: &size},
+				{Key: aws.String("hls/r-20260815T000000Z/seg-3.m4s"), Size: &size},
+				{Key: aws.String("hls/playlist.m3u8"), Size: &size},
+				{Key: aws.String("hls/playlist.m3u8.tmp"), Size: &size},
 			},
 			IsTruncated: aws.Bool(false),
 		}, nil
@@ -654,23 +743,101 @@ func TestCleanupOldSegments(t *testing.T) {
 		return &s3.DeleteObjectsOutput{}, nil
 	}
 
-	n, err := cleanupOldSegments(context.Background())
+	n, err := cleanupOldRuns(context.Background())
 	if err != nil {
-		t.Fatalf("cleanupOldSegments error: %v", err)
+		t.Fatalf("cleanupOldRuns error: %v", err)
 	}
+	if n != 1 {
+		t.Fatalf("deleted = %d, want 1 (oldest run)", n)
+	}
+
+	joined := strings.Join(deletedKeys, ",")
+	if !strings.Contains(joined, "hls/r-20260813T000000Z/seg-1.m4s") {
+		t.Fatalf("expected oldest run deleted, got: %v", deletedKeys)
+	}
+	if strings.Contains(joined, "playlist.m3u8") {
+		t.Fatalf("manifest must not be deleted: %v", deletedKeys)
+	}
+	if r2RunsRetained != 2 {
+		t.Fatalf("r2RunsRetained = %d, want 2", r2RunsRetained)
+	}
+	if r2BytesRetained != 2*size {
+		t.Fatalf("r2BytesRetained = %d, want %d", r2BytesRetained, 2*size)
+	}
+	if r2CleanupObjectsTotal != 1 {
+		t.Fatalf("r2CleanupObjectsTotal = %d, want 1", r2CleanupObjectsTotal)
+	}
+}
+
+func TestCleanupOldRuns_SizeCap(t *testing.T) {
+	prevClient := r2Client
+	prevBucket := bucket
+	prevList := listObjects
+	prevDelete := deleteObjects
+	prevTotal := r2CleanupObjectsTotal
+	prevRuns := r2RunsRetained
+	prevBytes := r2BytesRetained
+	defer func() {
+		r2Client = prevClient
+		bucket = prevBucket
+		listObjects = prevList
+		deleteObjects = prevDelete
+		r2CleanupObjectsTotal = prevTotal
+		r2RunsRetained = prevRuns
+		r2BytesRetained = prevBytes
+	}()
+
+	bucket = "linearvoid-media"
+	r2Client = &s3.Client{}
+
+	// Force a small cap so the size policy trims even the last 2 runs.
+	os.Setenv("R2_KEEP_RUNS", "2")
+	os.Setenv("R2_MAX_RETENTION_BYTES", "1024")
+	defer os.Unsetenv("R2_KEEP_RUNS")
+	defer os.Unsetenv("R2_MAX_RETENTION_BYTES")
+
+	size := int64(1024)
+	listObjects = func(_ *s3.Client, input *s3.ListObjectsV2Input) (*s3.ListObjectsV2Output, error) {
+		return &s3.ListObjectsV2Output{
+			Contents: []types.Object{
+				{Key: aws.String("hls/r-20260813T000000Z/seg-1.m4s"), Size: &size},
+				{Key: aws.String("hls/r-20260814T000000Z/seg-2.m4s"), Size: &size},
+				{Key: aws.String("hls/r-20260815T000000Z/seg-3.m4s"), Size: &size},
+			},
+			IsTruncated: aws.Bool(false),
+		}, nil
+	}
+
+	var deletedKeys []string
+	deleteObjects = func(_ *s3.Client, input *s3.DeleteObjectsInput) (*s3.DeleteObjectsOutput, error) {
+		for _, obj := range input.Delete.Objects {
+			deletedKeys = append(deletedKeys, aws.ToString(obj.Key))
+		}
+		return &s3.DeleteObjectsOutput{}, nil
+	}
+
+	n, err := cleanupOldRuns(context.Background())
+	if err != nil {
+		t.Fatalf("cleanupOldRuns error: %v", err)
+	}
+	// 3 runs -> keep 2 -> oldest deleted by count; remaining 2 bytes (2048) exceed
+	// the 1024 cap -> the older of the kept two is deleted, leaving only the newest.
 	if n != 2 {
 		t.Fatalf("deleted = %d, want 2", n)
 	}
 
 	joined := strings.Join(deletedKeys, ",")
-	if !strings.Contains(joined, "hls/r-old/seg-1.m4s") || !strings.Contains(joined, "hls/r-old/init.mp4") {
-		t.Fatalf("unexpected deleted keys: %v", deletedKeys)
+	if !strings.Contains(joined, "hls/r-20260813T000000Z/seg-1.m4s") {
+		t.Fatalf("expected oldest run deleted, got: %v", deletedKeys)
 	}
-	if strings.Contains(joined, "playlist.m3u8") {
-		t.Fatalf("manifest must not be deleted: %v", deletedKeys)
+	if !strings.Contains(joined, "hls/r-20260814T000000Z/seg-2.m4s") {
+		t.Fatalf("expected second-oldest run deleted by size cap, got: %v", deletedKeys)
 	}
-	if r2CleanupObjectsTotal != 2 {
-		t.Fatalf("r2CleanupObjectsTotal = %d, want 2", r2CleanupObjectsTotal)
+	if strings.Contains(joined, "hls/r-20260815T000000Z/seg-3.m4s") {
+		t.Fatalf("newest run must be kept, got: %v", deletedKeys)
+	}
+	if r2RunsRetained != 1 {
+		t.Fatalf("r2RunsRetained = %d, want 1", r2RunsRetained)
 	}
 }
 

backend/playout/start-playout.sh

@@ -3,8 +3,8 @@ set -e
 
 echo "=== Linear Void Playout Service Starting ==="
 echo "Playlist path: ${PLAYLIST_PATH:=/app/playlist/playlist.txt}"
-echo "HLS segment duration: ${HLS_SEGMENT_DURATION:-6}s"
-echo "HLS window size: ${HLS_WINDOW_SIZE:-24}"
+echo "HLS segment duration: ${HLS_SEGMENT_DURATION:-2}s"
+echo "HLS window size: ${HLS_WINDOW_SIZE:-12}"
 
 # Ensure playlist directory and file exist
 mkdir -p "$(dirname "$PLAYLIST_PATH")"
@@ -17,15 +17,7 @@ file '/app/media/demo2.mp4'
 EOF
 fi
 
-# Run normalization validation using the Go tool
-echo "Validating playlist and enforcing normalization (ADR-001/ADR-003)..."
-if ! /app/playlist-manager -validate "$PLAYLIST_PATH"; then
-  echo "ERROR: Playlist validation failed. Check logs."
-  exit 1
-fi
-
-echo "Playlist validated successfully. Starting continuous HLS playout..."
-
-# Single source of truth: the Go binary runs the full loop (HLS, R2 sync,
-# slate fallback, health server). The shell no longer duplicates the FFmpeg loop.
+# Single source of truth: the Go binary validates the playlist (ADR-001/ADR-003
+# normalization enforcement) and runs the full loop (HLS, R2 sync, slate
+# fallback, health server). Validation runs once inside the binary.
 exec /app/playlist-manager "$PLAYLIST_PATH"

docker-compose.prod.yml

@@ -11,6 +11,8 @@ services:
     restart: unless-stopped
     env_file:
       - .env
+    ports:
+      - "${PLAYOUT_PORT:-8081}:8081"
     volumes:
       - ./data/playlist:/app/playlist:rw
       - ./data/media:/app/media:ro
@@ -56,7 +58,7 @@ services:
     container_name: linearvoid-nginx
     restart: unless-stopped
     ports:
-      - "${NGINX_PORT:-8080}:80"
+      - "${NGINX_PORT:-8082}:80"
     volumes:
       - ./data/hls:/srv/hls:ro
     networks:

docker-compose.yml

@@ -13,6 +13,8 @@ services:
     restart: unless-stopped
     env_file:
       - .env
+    ports:
+      - "${PLAYOUT_PORT:-8081}:8081"
     volumes:
       - ./data/playlist:/app/playlist:ro
       - ./data/media:/app/media:ro
@@ -64,7 +66,7 @@ services:
     container_name: linearvoid-nginx
     restart: unless-stopped
     ports:
-      - "${NGINX_PORT:-8080}:80"
+      - "${NGINX_PORT:-8082}:80"
     volumes:
       - ./frontend/dist:/srv/frontend:ro
       - ./data/hls:/srv/hls:ro

docs/API.md

@@ -1,216 +1,101 @@
-# API Reference
+# Linear Void API Reference
 
-This document describes the currently implemented API surfaces in Linear Void.
+Base URLs: Public/admin app `https://linearvoid.tv` · WebSocket `wss://linearvoid.tv/ws` · Local fallback `http://localhost` / `ws://localhost/ws`
 
-## Base URLs
+## Authentication
 
-- Public app and admin: <https://linearvoid.tv>
-- WebSocket endpoint: <wss://linearvoid.tv/ws>
-- Local fallback: <http://localhost> and <ws://localhost/ws>
+Admin REST endpoints use a cookie session:
 
-## Authentication Model
-
-Admin REST endpoints use a cookie session.
-
-- Cookie name: linearvoid_admin_session
-- Cookie attributes: HttpOnly, SameSite=Strict, Secure when request is HTTPS (or X-Forwarded-Proto=https)
-- Session TTL: controlled by ADMIN_SESSION_TTL_HOURS (default 12h)
+- Cookie: `linearvoid_admin_session` — `HttpOnly`, `SameSite=Strict`, `Secure` when request is HTTPS (or `X-Forwarded-Proto=https`)
+- Session TTL: `ADMIN_SESSION_TTL_HOURS` (default 12h)
 
 ## Chat WebSocket API
 
-### Connect
+Connect: `GET /ws?username={name}&room={room}` — client/server exchange JSON messages.
 
-Endpoint:
+| Message (client→server) | Field                                                                            | Notes |
+| ----------------------- | -------------------------------------------------------------------------------- | ----- |
+| `type`                  | `public` for room chat, `private` for DM routing; `system` for moderation/events |       |
+| `room`                  | `main`; private DMs use `dm:userA-userB` (usernames sorted)                      |       |
+| `username`              | sender                                                                           |       |
+| `text`                  | message body                                                                     |       |
 
-- GET /ws?username={name}&room={room}
-
-Client sends JSON messages over the socket.
-
-### Message Envelope (client -> server)
+Server→client envelope adds `"timestamp"` (ISO-8601, e.g. `2026-08-04T12:34:56Z`).
 
 ```json
-{
-  "room": "main",
-  "username": "viewer123",
-  "text": "hello world",
-  "type": "public"
-}
-```
-
-Notes:
-
-- type is usually public for room chat and private for DM room routing.
-- Private room names are formatted like dm:userA-userB (sorted usernames).
-- Moderation/system events are sent as type=system messages.
-
-### Message Envelope (server -> client)
-
-```json
-{
-  "room": "main",
-  "username": "viewer123",
-  "text": "hello world",
-  "type": "public",
-  "timestamp": "2026-08-04T12:34:56Z"
-}
+{ "room": "main", "username": "viewer123", "text": "hello world", "type": "public" }
 ```
 
 ## Admin Session API
 
-### Login
+| Method & Path            | Auth                                | Description                   |
+| ------------------------ | ----------------------------------- | ----------------------------- |
+| `POST /api/admin/login`  | none                                | Authenticate                  |
+| `POST /api/admin/logout` | optional (clears cookie either way) | Logout                        |
+| `GET /api/admin/session` | cookie required                     | Session check; 401 if invalid |
 
-- POST /api/admin/login
-- Auth: none
-
-Request:
+Login request:
 
 ```json
-{
-  "username": "admin",
-  "password": "your-password"
-}
+{ "username": "admin", "password": "your-password" }
 ```
 
-Success response:
+Login/logout/session success response (session returns 401 on failure):
 
 ```json
-{
-  "authenticated": true,
-  "username": "admin"
-}
+{ "authenticated": true, "username": "admin" }
 ```
 
-### Logout
+Logout response: `{ "authenticated": false }`
 
-- POST /api/admin/logout
-- Auth: optional (clears cookie either way)
+## Admin Data API (auth required)
 
-Response:
+| Method & Path                             | Returns                                                                                              |
+| ----------------------------------------- | ---------------------------------------------------------------------------------------------------- |
+| `GET /api/admin/overview`                 | Aggregate counters + runtime/system snapshot                                                         |
+| `GET /api/admin/playlist`                 | Playlist                                                                                             |
+| `PUT /api/admin/playlist`                 | Set playlist (body: `{ "content": "file '/app/media/video1.mp4'\nfile '/app/media/video2.mp4'\n" }`) |
+| `GET /api/admin/schedule`                 | Parsed playlist entries + matched media metadata/runtime                                             |
+| `GET /api/admin/moderation/logs?limit=50` | Moderation logs                                                                                      |
+| `POST /api/admin/moderation/actions`      | Perform moderation action                                                                            |
+| `GET /api/admin/analytics`                | Totals (messages, moderation actions, rooms) + top room summary                                      |
+| `GET /api/admin/system`                   | Health snapshot (chat/playout/redis/database) + config metadata                                      |
+| `GET /api/admin/users`                    | Presence users (room + username) when Redis available                                                |
+
+Moderation action request:
 
 ```json
-{
-  "authenticated": false
-}
+{ "action": "mute", "target": "username-or-id", "moderator": "admin", "reason": "spam", "room": "main" }
 ```
 
-### Session Check
-
-- GET /api/admin/session
-- Auth: cookie required
-
-Success response:
-
-```json
-{
-  "authenticated": true,
-  "username": "admin"
-}
-```
-
-Failure:
-
-- 401 unauthorized
-
-## Admin Data API
-
-All endpoints below require admin authentication.
-
-### Overview
-
-- GET /api/admin/overview
-
-Returns aggregate counters and runtime/system snapshot fields.
-
-### Playlist
-
-- GET /api/admin/playlist
-- PUT /api/admin/playlist
-
-PUT request:
-
-```json
-{
-  "content": "file '/app/media/video1.mp4'\nfile '/app/media/video2.mp4'\n"
-}
-```
-
-### Schedule
-
-- GET /api/admin/schedule
-
-Returns parsed playlist entries plus matched media metadata/runtime.
-
-### Moderation
-
-- GET /api/admin/moderation/logs?limit=50
-- POST /api/admin/moderation/actions
-
-Action request:
-
-```json
-{
-  "action": "mute",
-  "target": "username-or-id",
-  "moderator": "admin",
-  "reason": "spam",
-  "room": "main"
-}
-```
-
-### Analytics
-
-- GET /api/admin/analytics
-
-Returns totals (messages, moderation actions, rooms) and top room summary.
-
-### System
-
-- GET /api/admin/system
-
-Returns service health snapshot for chat/playout/redis/database and config metadata.
-
-### Active Users
-
-- GET /api/admin/users
-
-Returns currently discovered presence users (room + username) when Redis is available.
-
 ## Media Library API
 
-### Media Collection
+| Method & Path                  | Description                                       |
+| ------------------------------ | ------------------------------------------------- |
+| `GET /api/admin/media`         | List media                                        |
+| `POST /api/admin/media`        | Multipart upload (legacy), form field name `file` |
+| `GET /api/admin/media/{id}`    | Get item                                          |
+| `PUT /api/admin/media/{id}`    | Update item                                       |
+| `DELETE /api/admin/media/{id}` | Delete item                                       |
 
-- GET /api/admin/media
-- POST /api/admin/media (multipart upload, legacy path)
-
-Multipart form field name: file
-
-### Media Item
-
-- GET /api/admin/media/{id}
-- PUT /api/admin/media/{id}
-- DELETE /api/admin/media/{id}
-
-PUT request:
+PUT media body:
 
 ```json
-{
-  "status": "ready",
-  "title": "My Clip",
-  "tags": "music,live"
-}
+{ "status": "ready", "title": "My Clip", "tags": "music,live" }
 ```
 
 ## Resumable Upload API
 
-Maximum file size: 2GB
+Max file size: **2GB** · Default chunk size: **8MB**
 
-Default chunk size: 8MB
+| Method & Path                                                      | Description                                                       |
+| ------------------------------------------------------------------ | ----------------------------------------------------------------- |
+| `POST /api/admin/media/uploads`                                    | Initialize or resume session                                      |
+| `GET /api/admin/media/uploads/{uploadId}`                          | Check upload status                                               |
+| `PUT /api/admin/media/uploads/{uploadId}/chunk?index={chunkIndex}` | Upload chunk (`Content-Type: application/octet-stream`)           |
+| `POST /api/admin/media/uploads/{uploadId}/complete`                | Complete upload → returns media listing (`files` + `generatedAt`) |
 
-### Initialize or Resume Session
-
-- POST /api/admin/media/uploads
-
-Request:
+Initialize request:
 
 ```json
 {
@@ -223,7 +108,7 @@ Request:
 }
 ```
 
-Response:
+Status envelope (returned by init, status check, and chunk upload):
 
 ```json
 {
@@ -237,61 +122,33 @@ Response:
 }
 ```
 
-### Check Upload Status
+## Metrics & Health
 
-- GET /api/admin/media/uploads/{uploadId}
-
-Returns same status envelope as initialize.
-
-### Upload Chunk
-
-- PUT /api/admin/media/uploads/{uploadId}/chunk?index={chunkIndex}
-- Content-Type: application/octet-stream
-
-Response returns updated upload status envelope.
-
-### Complete Upload
-
-- POST /api/admin/media/uploads/{uploadId}/complete
-
-On success returns admin media listing payload (files + generatedAt).
-
-## Metrics and Health
-
-### Chat metrics
-
-- GET /metrics
-
-Prometheus exposition includes connected clients, stored messages, moderation actions, Redis indicators, active admin sessions, and build info.
-
-### Health
-
-- GET /health on chat service
-- GET /health on playout service
-- GET /health on nginx reverse proxy
+- `GET /metrics` (chat) — Prometheus: connected clients, stored messages, moderation actions, Redis indicators, active admin sessions, build info
+- `GET /health` — chat service, playout service, and nginx reverse proxy each expose it
 
 ## HTTP Status Conventions
 
 | Code | Meaning               | When                                                        |
 | ---- | --------------------- | ----------------------------------------------------------- |
-| 200  | OK                    | Successful GET, PUT, or action                              |
+| 200  | OK                    | Successful GET, PUT, action                                 |
 | 201  | Created               | Successful POST (upload init, override created)             |
 | 204  | No Content            | Successful DELETE, CORS preflight                           |
 | 400  | Bad Request           | Malformed JSON, missing required fields, invalid input      |
-| 401  | Unauthorized          | Missing or expired admin session                            |
-| 403  | Forbidden             | Valid session but insufficient permissions                  |
+| 401  | Unauthorized          | Missing/expired admin session                               |
+| 403  | Forbidden             | Valid session, insufficient permissions                     |
 | 404  | Not Found             | Resource doesn't exist                                      |
-| 405  | Method Not Allowed    | Wrong HTTP method for endpoint                              |
+| 405  | Method Not Allowed    | Wrong method for endpoint                                   |
 | 409  | Conflict              | Schedule override overlap, upload ID mismatch               |
 | 413  | Payload Too Large     | Upload exceeds 2 GB limit                                   |
 | 422  | Unprocessable         | File fails normalization check (codec, fps, resolution)     |
-| 429  | Too Many Requests     | Rate limit exceeded (chat messages, WS connections, login)  |
+| 429  | Too Many Requests     | Rate limit exceeded (chat, WS, login)                       |
 | 500  | Internal Server Error | Unexpected server failure                                   |
 | 503  | Service Unavailable   | Redis unavailable for user actions, playout health degraded |
 
 ## Rate Limiting
 
-All limits are per-IP (from `X-Forwarded-For` or `RemoteAddr`) unless noted.
+Per-IP (from `X-Forwarded-For` or `RemoteAddr`) unless noted.
 
 | Resource              | Limit                              | Window                                   | Response                                               |
 | --------------------- | ---------------------------------- | ---------------------------------------- | ------------------------------------------------------ |
@@ -299,9 +156,4 @@ All limits are per-IP (from `X-Forwarded-For` or `RemoteAddr`) unless noted.
 | WebSocket connections | 5                                  | 60s                                      | 429 + `"too many connections from this IP"`            |
 | Admin login           | 5                                  | 5min                                     | 429 + `"too many login attempts, try again later"`     |
 
-Rate limit headers on applicable responses:
-
-- `X-RateLimit-Limit`: maximum requests per window
-- `Retry-After`: seconds until window resets
-- 413: payload too large
-- 500: internal server error
+Rate limit headers (applicable responses): `X-RateLimit-Limit` (max requests/window), `Retry-After` (seconds until reset). 413 = payload too large; 500 = internal server error.

docs/COOLIFY.md

@@ -0,0 +1,148 @@
+# Coolify Deployment
+
+*Last Updated: 2026-08-28*
+
+Alternative deployment path for Linear Void using [Coolify](https://github.com/coollabsio/coolify/tree/v4.3.14) (v4.3.14), an open-source self-hostable PaaS that manages servers, apps, and databases over SSH. Drop-in alternative to the flows in [SETUP.md](./SETUP.md): local dev (`docker-compose.yml`) and test-lab/prod (`docker-compose.prod.yml` via Gitea Actions CD). Coolify deploys the same Compose stack through a web UI and its own Traefik reverse proxy for automatic HTTPS, domain routing, and one-click redeploys (no manual `DOCKER_HOST`).
+
+## Why Coolify
+
+- **Single control plane**: manage playout, chat, nginx, redis, db (plus exporters/monitoring) from one dashboard.
+- **Automatic TLS**: Let's Encrypt provisioning/renewal per app/domain, replacing the manual Cloudflare + nginx origin TLS setup for the app hostname.
+- **No vendor lock-in**: config lives on your server; containers keep running if you stop using Coolify — you only lose automation.
+- **Simpler cutover**: on Hetzner VPS provisioning, point Coolify at the new server and redeploy instead of editing CD `DOCKER_HOST`.
+
+> Replaces the *deployment orchestration* layer only. Cloudflare DNS, R2 media bucket, cache rules, and the edge Worker from [SETUP.md](./SETUP.md) are unchanged.
+
+## Prerequisites
+
+- Linux server with **Docker Engine** (any supported distro).
+- **SSH access** (root or passwordless `sudo`) from the Coolify install machine — Coolify provisions the target over SSH.
+- **2 GB RAM / 20 GB disk** minimum for Coolify itself (plus resources for the stack).
+- A domain pointed at the server for HTTPS (e.g. `linearvoid.tv`).
+- Ports `80` and `443` reachable for Traefik.
+- Gitea registry credentials (`CI_REGISTRY_USER` / `CI_REGISTRY_TOKEN`) if pulling prebuilt `playout`, `chat`, `nginx` images.
+
+## Installation
+
+Recommended on a dedicated server:
+
+```bash
+curl -fsSL https://cdn.coollabs.io/coolify/install.sh | bash
+```
+
+The installer boots a Docker-based Coolify instance and prints the URL and default credentials. Complete first-run setup in the dashboard. For non-interactive/offline installs, see the [installation docs](https://coolify.io/docs/installation) and [install script source](https://github.com/coollabsio/coolify/blob/v4.3.14/scripts/install.sh).
+
+## Add the Linear Void server
+
+1. **Servers → Add**.
+2. Choose your provider (or **Localhost** if Coolify runs on the same host as the stack).
+3. Provide SSH details; for a remote host Coolify validates and installs its agent.
+4. Confirm the server shows **Reachable** before creating resources.
+
+### Current topology (verified 2026-08-30)
+
+Active test-lab setup is **two-tier**:
+
+| Component | Host | Exposes | Docker TCP | Notes |
+|---|---|---|---|---|
+| **Coolify dashboard** | `192.168.88.18` | `https://coolify.allucanget.biz` via Cloudflare tunnel (`coolify-cloudflared`); dashboard at `8000` | No | Manages target servers over SSH |
+| **Deployment target** | `192.168.88.40` | `coolify-proxy` = Traefik on 80/443; port `8000` bound by Portainer, not Coolify | Yes | Per-server helpers (`coolify-proxy`, `coolify-sentinel`, `coolify-cloudflared`) + the stack via `docker-compose.prod.yml` (`linearvoid` compose project, 11 containers) |
+
+The main `coolify` app container runs only on the dashboard host `192.168.88.18`; Compose-resource services land on the target `192.168.88.40`. The dashboard host exposes no Docker TCP socket, so run the checks below against the **target** `192.168.88.40`:
+
+```bash
+./scripts/check-env.sh      # env/deployment health on target (containers, helpers, health)
+./scripts/check-tokens.sh   # Gitea registry, Cloudflare, R2, admin credential access
+```
+
+## Deploy the stack
+
+Deploy as a **Docker Compose** resource (not individual apps):
+
+1. **Projects → Add → Docker Compose**.
+2. Select the target server.
+3. In **Build Settings**, set the compose file source:
+   - **Git Repository** → point at `https://git.allucanget.biz/zwitschi/linearvoid.git`, compose file `docker-compose.prod.yml` (Coolify pulls on each redeploy).
+   - **Raw Compose** → paste `docker-compose.prod.yml` directly to manage manually.
+4. For the Git source, add the Gitea registry login as a **Deployment Key / environment secret** so `playout`, `chat`, `nginx` images can be pulled. Alternatively switch those services to `build:` blocks (as in `docker-compose.yml`) to build from source on the server.
+
+### Environment variables
+
+Map Linear Void's `.env` values into Coolify's **Environment Variables** tab:
+
+| Variable | Purpose |
+|---|---|
+| `DATABASE_URL` | `postgres://linearvoid:<password>@db:5432/linearvoid?sslmode=disable` |
+| `REDIS_URL` | `redis://redis:6379/0` |
+| `POSTGRES_PASSWORD` | DB password (used by `db` service) |
+| `ADMIN_USERNAME` / `ADMIN_PASSWORD` | Admin panel credentials |
+| `ADMIN_SESSION_TTL_HOURS` | Admin session lifetime |
+| `R2_ACCESS_KEY_ID` / `R2_SECRET_ACCESS_KEY` | Cloudflare R2 credentials |
+| `R2_BUCKET_NAME` / `R2_ENDPOINT` / `R2_REGION` | R2 bucket config (`R2_REGION=auto`) |
+| `MEDIA_DOMAIN` | `media.linearvoid.tv` |
+| `VITE_PUBLIC_APP_URL` | `https://linearvoid.tv` |
+| `VITE_PUBLIC_WS_URL` | `wss://linearvoid.tv/ws` |
+| `VITE_PUBLIC_HLS_URL` | `https://media.linearvoid.tv/hls/playlist.m3u8` |
+
+Keep internal service URLs on compose hostnames (`redis`, `db`) exactly as in `docker-compose.prod.yml`.
+
+### Persistent storage
+
+Named volumes (`redis-data`, `postgres-data`) and bind mounts under `./data` must be configured under **Persistent Storage** to survive redeploys:
+
+| Mount | Purpose |
+|---|---|
+| `data/playlist` | Playout playlist file |
+| `data/media` | Normalized source videos (read-only to playout) |
+| `data/hls` | Generated HLS manifests/segments |
+| `data/chat` | Chat service working data |
+| `data/slate` | Slate media |
+| `redis-data` | Redis append-only persistence |
+| `postgres-data` | PostgreSQL data directory |
+
+If Coolify doesn't auto-create the dirs, create them before first deploy:
+
+```bash
+mkdir -p data/playlist data/media data/hls data/chat
+```
+
+### Domains and proxy
+
+1. Assign the **nginx** service a public domain (e.g. `linearvoid.tv`); Traefik terminates TLS and forwards to `nginx`.
+2. Set **Ports Exposes** so Traefik targets the nginx port (`80` inside container, mapped from `${NGINX_PORT:-8082}:80`).
+3. Do **not** publicly expose playout (`8081`) or chat (`3000`) health ports — keep them internal and route through nginx.
+4. For `media.linearvoid.tv`, keep serving HLS from the R2 custom domain (per SETUP.md) rather than routing through Coolify, so segments stay on Cloudflare's edge cache.
+
+## Post-deploy verification
+
+- `docker compose ps` on the server shows all services healthy.
+- Playout health: `https://<host>/health` via nginx (internal `:8081/health`).
+- Chat health: internal `:3000/health`.
+- HLS written to `data/hls` and served from `media.linearvoid.tv`.
+- Chat WebSocket connects over `wss://linearvoid.tv/ws`.
+- Admin login works at `/admin` with configured credentials.
+
+## Monitoring
+
+Coolify exposes container logs/basic metrics per resource. The stack ships Prometheus + Grafana (`monitoring/`) plus nginx/redis exporters. To reach them via Coolify, expose Grafana on a restricted subdomain (e.g. `grafana.linearvoid.tv`) and keep Prometheus internal, mirroring SETUP.md access guidance.
+
+## Cutover from Gitea CD
+
+1. Deploy the stack as a Coolify compose resource (above).
+2. Point Cloudflare DNS for `linearvoid.tv` / `www` at the server IP (proxied).
+3. Disable (or keep as fallback) the `cd` job in the Gitea workflow so the two orchestrators don't fight over the same `DOCKER_HOST`.
+4. Keep the Gitea **CI** (build + push images) job so Coolify keeps pulling `:latest` from the registry.
+
+## Troubleshooting
+
+- **Image pull fails**: set Gitea registry credentials in the compose secret store, or switch services to `build:` blocks.
+- **Healthchecks fail**: confirm `curl` exists in the built images (healthchecks rely on it).
+- **TLS errors**: check the domain's DNS points at the Coolify server and ports `80`/`443` are open.
+- **HLS not updating**: verify `data/hls` is writable and playout restarts cleanly; see SETUP.md normalization troubleshooting.
+- **Ports conflict**: Coolify's Traefik binds `80`/`443`; do not also map those to the host in the compose file.
+
+## Further reading
+
+- Coolify repo (v4.3.14): <https://github.com/coollabsio/coolify/tree/v4.3.14>
+- Coolify docs: <https://coolify.io/docs>
+- [DEPLOYMENT.md](./DEPLOYMENT.md) · [SETUP.md](./SETUP.md)

docs/DEPLOYMENT.md

@@ -1,63 +1,28 @@
 # Deployment Guide
 
-## Table of Contents
-
-- [Deployment Guide](#deployment-guide)
-  - [Table of Contents](#table-of-contents)
-  - [Environments](#environments)
-  - [Production Target (Hetzner)](#production-target-hetzner)
-    - [Planned Infrastructure](#planned-infrastructure)
-    - [Cutover Plan](#cutover-plan)
-    - [Production Preflight](#production-preflight)
-    - [Firewall / Security Group](#firewall--security-group)
-    - [Pipeline Readiness](#pipeline-readiness)
-  - [Target Topology](#target-topology)
-  - [Prerequisites](#prerequisites)
-  - [Environment File](#environment-file)
-  - [Pull-and-Run Deployment](#pull-and-run-deployment)
-  - [First-Time or Recovery Provisioning](#first-time-or-recovery-provisioning)
-  - [Health Verification](#health-verification)
-  - [Zero/Low-Downtime Update Pattern](#zerolow-downtime-update-pattern)
-  - [Rollback](#rollback)
-  - [Operational Tasks](#operational-tasks)
-    - [View logs](#view-logs)
-    - [Restart one service](#restart-one-service)
-    - [Stop stack](#stop-stack)
-    - [Keep volumes (recommended) vs remove volumes (destructive)](#keep-volumes-recommended-vs-remove-volumes-destructive)
-  - [Nginx and Proxy Considerations](#nginx-and-proxy-considerations)
-  - [Security Baseline](#security-baseline)
-  - [Common Deployment Pitfalls](#common-deployment-pitfalls)
-  - [Troubleshooting](#troubleshooting)
-
 ## Environments
 
-Linear Void runs in two separate environments:
-
 | Environment    | Host                       | Docker Target              | Purpose                                        |
 | -------------- | -------------------------- | -------------------------- | ---------------------------------------------- |
 | **Test Lab**   | `192.168.88.40` (internal) | `tcp://192.168.88.40:2375` | Integration testing, staging, CI/CD validation |
-| **Production** | Hetzner VPS (TBD)          | `tcp://<hetzner-ip>:2375`  | Public linearvoid.tv serving real viewers      |
+| **Production** | Hetzner VPS (TBD)          | `tcp://<hetzner-ip>:2375`  | Public linearvoid.tv                           |
 
-**Current state (2026-08-06):** Test lab is active and receives automated CD deploys. Production Hetzner VPS is **not yet provisioned**. The CD workflow (`cd.yml`) and `docker-compose.prod.yml` currently target the test lab. Switching to production requires only changing `DOCKER_HOST` — see [Production Deployment](#production-target-hetzner) below.
+**Current state (2026-08-06):** Test lab is active and receives automated CD deploys. Production Hetzner VPS is **not yet provisioned**. `cd.yml` and `docker-compose.prod.yml` target the test lab; switching to production only requires changing `DOCKER_HOST`.
 
 ## Production Target (Hetzner)
 
-**Status: Not provisioned.** The production deployment plan is documented and the pipeline is ready — only the VPS needs to be created.
+**Status: Not provisioned.** Plan documented and pipeline ready; only the VPS needs creation.
 
 ### Planned Infrastructure
 
-| Resource       | Spec                                             |
-| -------------- | ------------------------------------------------ |
-| Provider       | Hetzner Cloud                                    |
-| Region         | Ashburn, VA (us-east)                            |
-| Instance       | CX22 (2 vCPU, 4 GB RAM, 40 GB SSD) or equivalent |
-| OS             | Ubuntu 24.04 LTS or Debian 12                    |
-| Bandwidth      | 20 TB included traffic/month                     |
-| Estimated cost | ~€7–10/month base                                |
+- Provider: Hetzner Cloud, Region: Ashburn, VA (us-east)
+- Instance: CX22 (2 vCPU, 4 GB RAM, 40 GB SSD) or equivalent
+- OS: Ubuntu 24.04 LTS or Debian 12
+- Bandwidth: 20 TB included traffic/month; Estimated cost: ~€7–10/month base
 
 ### Cutover Plan
 
-1. Provision Hetzner VPS (see [Production Deployment](#production-target-hetzner) section)
+1. Provision Hetzner VPS
 2. Add production secrets to Gitea (`CI_PROD_*` prefix)
 3. Update `DOCKER_HOST` in CD workflow to Hetzner public IP
 4. Trigger deploy → validate health checks
@@ -65,18 +30,14 @@ Linear Void runs in two separate environments:
 6. Enable HTTPS, WAF, cache rules on production zone
 7. Decommission test lab as public target
 
-### Production Preflight
+### Production Preflight (minimum readiness gate)
 
-Before the cutover in the plan above, treat the following as the minimum readiness gate:
-
-1. Verify the Hetzner host is reachable over SSH and Docker remote API TLS is working for the Gitea runner.
-2. Confirm all required production secrets are present under the `CI_PROD_*` prefix.
-3. Confirm `linearvoid.tv`, `www.linearvoid.tv`, and `media.linearvoid.tv` DNS targets are correct for the intended phase.
-4. Complete the manual Cloudflare/UptimeRobot checklist and attach evidence in [Cloudflare Uptimerobot checklist](../docs/operations/cloudflare-uptimerobot-checklist.md).
-5. Run the production smoke path end to end: viewer SPA, HLS playback, chat connect, admin login, upload, playlist save, and playout health.
-6. Verify rollback is still possible before the final DNS switch.
-
-Readiness is not complete until the checklist above and the manual evidence tracker are both satisfied.
+1. Verify Hetzner host reachable over SSH and Docker remote API TLS works for the Gitea runner
+2. Confirm all production secrets exist under `CI_PROD_*` prefix
+3. Confirm `linearvoid.tv`, `www.linearvoid.tv`, `media.linearvoid.tv` DNS targets match intended phase
+4. Complete manual Cloudflare/UptimeRobot checklist (see `docs/operations/cloudflare-uptimerobot-checklist.md`)
+5. Run production smoke path end-to-end: viewer SPA, HLS playback, chat connect, admin login, upload, playlist save, playout health
+6. Verify rollback possible before final DNS switch
 
 ### Firewall / Security Group
 
@@ -91,18 +52,14 @@ Readiness is not complete until the checklist above and the manual evidence trac
 
 ### Pipeline Readiness
 
-Three Gitea Actions workflows drive the full pipeline:
+| Workflow             | Trigger           | Purpose                                                                           |
+| -------------------- | ----------------- | --------------------------------------------------------------------------------- |
+| `ci.yml`             | Push to `main`    | Lint, build frontend, vet Go, build & push images (`:latest`, `:main`)            |
+| `cd.yml`             | Push to `main`    | Deploy to test lab (`192.168.88.40:2375`) — runs **in parallel** with CI          |
+| `release.yml`        | Tag push `*.*.*`  | Build & push versioned images (`:X.Y.Z`, `:X.Y`, `:latest`), create Gitea Release |
+| `workers-deploy.yml` | After CI succeeds | Deploy frontend to Cloudflare Workers                                             |
 
-| Workflow           | Trigger          | Purpose                                                                                  |
-| ------------------ | ---------------- | ---------------------------------------------------------------------------------------- |
-| `ci.yml`           | Push to `main`   | Lint, build frontend, vet Go, build & push Docker images (`:latest`, `:main`)            |
-| `cd.yml`           | Push to `main`   | Deploy to test lab (`192.168.88.40:2375`) — runs **in parallel** with CI                 |
-| `release.yml`      | Tag push `*.*.*` | Build & push versioned Docker images (`:X.Y.Z`, `:X.Y`, `:latest`), create Gitea Release |
-| `pages-deploy.yml` | Push to `main`   | Deploy frontend to Cloudflare Pages (see `docs/operations/gitea-cloudflare-pages.md`)    |
-
-CI and CD run simultaneously on every push to `main`. The deploy pulls whatever `:latest` images are available in the registry at that moment.
-
-The CD workflow (`cd.yml`) uses `DOCKER_HOST` to target the deployment host. To switch targets:
+CI and CD run simultaneously on every push to `main`. Deploy pulls whatever `:latest` images are available at that moment. `cd.yml` uses `DOCKER_HOST` to target the host:
 
 ```bash
 # Test lab (current)
@@ -112,7 +69,7 @@ DOCKER_HOST=tcp://192.168.88.40:2375
 DOCKER_HOST=tcp://<hetzner-public-ip>:2375
 ```
 
-No code changes needed in the pipeline. See `DEPLOYMENT.md` Production Deployment section for full checklist.
+No pipeline code changes needed to switch targets.
 
 ## Target Topology
 
@@ -126,12 +83,12 @@ No code changes needed in the pipeline. See `DEPLOYMENT.md` Production Deploymen
 
 - Docker Engine + Docker Compose plugin
 - Network access to git.allucanget.biz
-- A valid .env file with production values
-- DNS configured for linearvoid.tv (and media.linearvoid.tv if split media domain is used)
+- Valid `.env` file with production values
+- DNS configured for `linearvoid.tv` (and `media.linearvoid.tv` if split media domain used)
 
 ## Environment File
 
-Use `.env` with at least:
+`.env` requires at minimum:
 
 - `DATABASE_URL` — PostgreSQL connection string
 - `REDIS_URL` — Redis connection string
@@ -141,12 +98,12 @@ Use `.env` with at least:
 
 **Cloudflare R2 (required for HLS segment sync to edge):**
 
-- `R2_ACCESS_KEY_ID` / `R2_SECRET_ACCESS_KEY` — R2 write credentials
-- `R2_BUCKET_NAME` — bucket name (e.g. `linearvoid-media`)
+- `R2_ACCESS_KEY_ID` / `R2_SECRET_ACCESS_KEY`
+- `R2_BUCKET_NAME` (e.g. `linearvoid-media`)
 - `R2_ENDPOINT` — S3-compatible endpoint URL
 - `R2_REGION` — set to `auto`
 
-**Cloudflare (optional, for Workers/Pages):**
+**Cloudflare (optional, for Workers):**
 
 - `CLOUDFLARE_API_TOKEN` / `CLOUDFLARE_ACCOUNT_ID`
 - `MEDIA_DOMAIN` — split media domain (e.g. `media.linearvoid.tv`)
@@ -155,149 +112,99 @@ See `.env.example` and `SETUP.md` for full variable reference.
 
 ### Gitea Secrets
 
-The CD workflow (`cd.yml`) injects sensitive values from Gitea secrets into the `.env` file at deploy time. The following secrets must be configured in the Gitea repository:
+`cd.yml` injects sensitive values from Gitea secrets into `.env` at deploy time. Configure in the Gitea repository:
 
-| Secret Name               | Description                          |
-| ------------------------- | ------------------------------------ |
-| `DATABASE_URL`            | PostgreSQL connection string         |
-| `ADMIN_USERNAME`          | Admin dashboard username             |
-| `ADMIN_PASSWORD`          | Admin dashboard password             |
-| `R2_ACCESS_KEY_ID`        | Cloudflare R2 access key             |
-| `R2_SECRET_ACCESS_KEY`    | Cloudflare R2 secret key             |
-| `R2_BUCKET_NAME`          | R2 bucket name                       |
-| `R2_ENDPOINT`             | R2 S3-compatible endpoint URL        |
-| `R2_ADMIN_ACCESS_KEY_ID`  | R2 admin access key (optional)       |
-| `R2_ADMIN_SECRET_ACCESS_KEY` | R2 admin secret key (optional)    |
-| `R2_ADMIN_ENDPOINT`       | R2 admin endpoint (optional)         |
-| `CLOUDFLARE_API_TOKEN`    | Cloudflare API token (optional)      |
-| `CLOUDFLARE_ACCOUNT_ID`   | Cloudflare account ID (optional)     |
-| `CI_REGISTRY_USER`        | Gitea container registry username    |
-| `CI_REGISTRY_TOKEN`       | Gitea container registry token       |
+`DATABASE_URL`, `ADMIN_USERNAME`, `ADMIN_PASSWORD`, `R2_ACCESS_KEY_ID`, `R2_SECRET_ACCESS_KEY`, `R2_BUCKET_NAME`, `R2_ENDPOINT`, `R2_ADMIN_ACCESS_KEY_ID` (optional), `R2_ADMIN_SECRET_ACCESS_KEY` (optional), `R2_ADMIN_ENDPOINT` (optional), `CLOUDFLARE_API_TOKEN` (optional), `CLOUDFLARE_ACCOUNT_ID` (optional), `CI_REGISTRY_USER`, `CI_REGISTRY_TOKEN`
 
 ## Pull-and-Run Deployment
 
-Images are built and pushed automatically by the CI and release workflows (see [CONTRIBUTING.md](../CONTRIBUTING.md#versioning-and-releases) for the release procedure).
-
-Every push to `main` triggers CI (builds `:main` and `:latest` tags) and CD (deploys `:latest` to test lab) in parallel. Semantic version tags (`*.*.*`, e.g. `1.0.9`) trigger the release workflow which builds `:X.Y.Z`, `:X.Y`, and `:latest` tags and creates a Gitea Release.
-
-From repository root:
+Images are built/pushed automatically by CI and release workflows. Every push to `main` triggers CI (`:main`, `:latest`) and CD (deploys `:latest` to test lab) in parallel. Semantic version tags (`*.*.*`, e.g. `1.0.9`) trigger the release workflow (`:X.Y.Z`, `:X.Y`, `:latest` + Gitea Release).
 
 ```bash
 docker login git.allucanget.biz
-
 docker compose -f docker-compose.prod.yml pull
 docker compose -f docker-compose.prod.yml up -d --remove-orphans
 ```
 
 ## First-Time or Recovery Provisioning
 
-1. Ensure host paths exist:
-
 ```bash
 mkdir -p data/chat data/playlist data/media data/hls
-```
-
-1. Confirm volumes and networks:
-
-```bash
 docker volume ls
 docker network ls
-```
-
-1. Start stack:
-
-```bash
 docker compose -f docker-compose.prod.yml up -d
 ```
 
 ## Health Verification
 
-After startup, check service status:
-
 ```bash
 docker compose -f docker-compose.prod.yml ps
 ```
 
 Endpoint checks:
 
-- App: <http://{host}:8080/>
-- Nginx health: <http://{host}:8080/health>
-- Chat health: <http://{host}:3000/health>
-- Playout health: <http://{host}:8081/health>
-- Prometheus: <http://{host}:9090>
-- Grafana: <http://{host}:3001>
+- App: `http://{host}:8080/`
+- Nginx health: `http://{host}:8080/health`
+- Chat health: `http://{host}:3000/health`
+- Playout health: `http://{host}:8081/health`
+- Prometheus: `http://{host}:9090`
+- Grafana: `http://{host}:3001`
 
 Domain checks:
 
-- <https://linearvoid.tv>
-- <https://linearvoid.tv/admin>
-- <https://linearvoid.tv/metrics>
+- `https://linearvoid.tv`
+- `https://linearvoid.tv/admin`
+- `https://linearvoid.tv/metrics`
 
 ## Zero/Low-Downtime Update Pattern
 
 1. Pull newest images.
-1. Recreate changed services only:
+2. Recreate changed services only:
 
-```bash
-docker compose -f docker-compose.prod.yml pull
+   ```bash
+   docker compose -f docker-compose.prod.yml pull
+   docker compose -f docker-compose.prod.yml up -d nginx chat playout
+   ```
 
-docker compose -f docker-compose.prod.yml up -d nginx chat playout
-```
-
-1. Re-run health checks.
+3. Re-run health checks.
 
 ## Rollback
 
 If new images are unhealthy:
 
-1. Pin previous tags in docker-compose.prod.yml.
+1. Pin previous tags in `docker-compose.prod.yml`.
 2. Pull pinned tags.
-3. Restart impacted services.
+3. Restart impacted services:
 
 ```bash
 docker compose -f docker-compose.prod.yml pull
-
 docker compose -f docker-compose.prod.yml up -d nginx chat playout
 ```
 
 ## Operational Tasks
 
-### View logs
-
 ```bash
+# View logs
 docker compose -f docker-compose.prod.yml logs -f nginx chat playout
-```
 
-### Restart one service
-
-```bash
+# Restart one service
 docker compose -f docker-compose.prod.yml restart chat
-```
 
-### Stop stack
-
-```bash
-docker compose -f docker-compose.prod.yml down
-```
-
-### Keep volumes (recommended) vs remove volumes (destructive)
-
-```bash
-# keep data
+# Stop stack (keep data)
 docker compose -f docker-compose.prod.yml down
 
-# destructive (remove data)
+# Stop stack — destructive (remove data)
 docker compose -f docker-compose.prod.yml down -v
 ```
 
 ## Nginx and Proxy Considerations
 
-- If an external proxy (for example Nginx Proxy Manager) fronts the stack, align request body limits with internal nginx, see [nginx.conf](../infrastructure/nginx/nginx.conf).
-- For resumable admin uploads, ensure client_max_body_size supports chunk requests in both proxy layers.
+- If an external proxy (e.g. Nginx Proxy Manager) fronts the stack, align request body limits with internal nginx (see `infrastructure/nginx/nginx.conf`).
+- For resumable admin uploads, ensure `client_max_body_size` supports chunk requests in both proxy layers.
 
 ## Security Baseline
 
-- Use strong ADMIN_PASSWORD and rotate regularly.
+- Use a strong `ADMIN_PASSWORD` and rotate regularly.
 - Keep Docker host access restricted.
 - Prefer HTTPS/WSS through Cloudflare.
 - Restrict exposed ports if only proxy access is needed.
@@ -321,5 +228,4 @@ docker compose -f docker-compose.prod.yml down -v
 | Prometheus no targets      | `curl http://<host>:9090/targets`             | Verify container network, scrape config                       |
 | Grafana no data            | Check datasource                              | Verify Prometheus URL `http://prometheus:9090` provisioned    |
 
-- Proxy-layer body size limits returning 413 on upload.
-- Health checks failing before service warmup is complete.
+Additional notes: proxy-layer body size limits can return 413 on upload; health checks may fail before service warmup completes.

docs/MONITORING.md

@@ -52,7 +52,7 @@ Config: `monitoring/prometheus/prometheus.yml` (15s scrape interval)
 - **Linear Void app**: `monitoring/grafana/dashboard.json` — clients, messages, moderation, Redis, sessions, scrape health
 - **Prometheus ops**: `monitoring/grafana/dashboards/dashboard.json` — TSDB, CPU/memory, WAL, compactions, query latency
 
-## Metrics checking
+## Metrics Checking
 
 See detailed scripts for querying Prometheus scrape targets:
 
@@ -78,21 +78,13 @@ See detailed scripts for querying Prometheus scrape targets:
 | Prometheus targets down | Check container network: all must be on `linearvoid-net`             |
 | Alert not firing        | Check Prometheus → Alerts tab, verify `rule_files` in prometheus.yml |
 
-### Grafana shows no data
+**Grafana no data**: ensure datasource points to `http://prometheus:9090`; confirm Prometheus is healthy; check dashboard query labels match active jobs.
 
-- Ensure datasource points to <http://prometheus:9090>.
-- Confirm Prometheus is healthy.
-- Check dashboard query labels match active jobs.
-
-### High cardinality or query slowness
-
-- Reduce unneeded labels.
-- Limit expensive dashboard panels.
-- Tune retention and query windows.
+**High cardinality / slow queries**: reduce unneeded labels; limit expensive panels; tune retention and query windows.
 
 ## Operational Recommendations
 
 - Keep alert thresholds tied to actual user impact.
 - Review dashboards after each release that changes endpoints or metrics names.
 - Periodically test alert routing and incident response flow.
-- Metrics retention is set to **30 days** time-based plus a **15GB** size cap (`--storage.tsdb.retention.time=30d --storage.tsdb.retention.size=15GB` on the `prometheus` service in `docker-compose.yml` and `docker-compose.prod.yml`). Increase the size cap before extending the time window so the TSDB does not hit the disk limit first.
+- Retention: **30 days** time-based plus **15GB** size cap (`--storage.tsdb.retention.time=30d --storage.tsdb.retention.size=15GB` on the `prometheus` service in `docker-compose.yml` and `docker-compose.prod.yml`). Increase the size cap before extending the time window so the TSDB does not hit the disk limit first.

docs/README.md

@@ -9,7 +9,7 @@ This folder contains the documentation for the Linear Void project. The document
 ## Infrastructure
 
 - Cloudflare
-  - [Cache Rules](./infrastructure/cloudflare/cache-rules.md) - Cache rules for Cloudflare Workers and Pages.
+  - [Cache Rules](./infrastructure/cloudflare/cache-rules.md) - Cache rules for Cloudflare Workers.
   - [Account Troubleshooting](./infrastructure/cloudflare/cloudflare-account-troubleshooting.md) - Troubleshooting Cloudflare account issues.
 - Monitoring
   - [Prometheus CURL](./infrastructure/monitoring/prometheus_curl.md) - Prometheus metrics endpoint testing with curl and jq. Doocumentation of available metrics and their meanings.

docs/SETUP.md

@@ -2,545 +2,149 @@
 
 **Last Updated**: 2026-08-03
 
-This guide covers prerequisites, environment setup, running the application locally and in production, and troubleshooting. It references the [architecture documentation](./docs/architecture/README.md)(`docs/architecture/README.md`).
+Prerequisites, environment setup, and running the app locally and in production. Architecture: [`docs/architecture/README.md`](./architecture/README.md).
 
 ## Prerequisites
 
-- **Docker & Docker Compose**: Required for running the full stack (playout, chat, nginx, redis, db, prometheus, grafana). Install from [docker.com](https://www.docker.com).
-- **Node.js 22+**: Required for frontend development and production builds (`cd frontend && npm install && npm run build`).
-- **npm**: Installed with Node.js and used for the frontend dependency installation and Vite workflow.
-- **Go 1.25+**: Recommended when working on the Go services outside Docker (`backend/chat`, `backend/playout`).
-- **FFmpeg**: Used by the playout service for HLS generation. Install via `apt install ffmpeg` (Linux), Homebrew (macOS), or Winget (Windows).
-- **Git**: For cloning and version control.
-- **Cloudflare Account**: For R2 storage and CDN (free tier sufficient for most use cases).
-- **Gitea Account**: Required if you want to push images to the self-hosted registry or use the repository's CI/CD flow.
-  - Set up the required Actions secrets in Gitea at: [https://git.allucanget.biz/zwitschi/linearvoid/settings/actions/secrets](https://git.allucanget.biz/zwitschi/linearvoid/settings/actions/secrets)
-  - Add the following secrets:
-    - `CI_REGISTRY_USER`: Your Gitea username (e.g., `zwitschi`)
-    - `CI_REGISTRY_TOKEN`: A Gitea Personal Access Token with `write:packages` scope
-    - `ADMIN_USERNAME`: The administrator username for the production admin panel (e.g., `admin`)
-    - `ADMIN_PASSWORD`: A strong password for the production admin panel
-    - `DATABASE_URL`: The production database connection string (e.g., `postgres://linearvoid:<password>@db:5432/linearvoid?sslmode=disable`)
-- **Hetzner Cloud Account** (optional for production): For the US-East VPS (Ashburn, VA recommended per ADR-001).
+- **Docker + Docker Compose** — runs the full stack (playout, chat, nginx, redis, db, prometheus, grafana).
+- **Node.js 22+ / npm** — frontend dev and builds (`cd frontend && npm install && npm run build`).
+- **Go 1.25+** — working on the Go services outside Docker.
+- **FFmpeg** — playout HLS generation (`apt install ffmpeg`).
+- **Git**.
+- **Cloudflare account** — R2 storage + CDN.
+- **Gitea account** — image registry + CI/CD. Set Actions secrets at `https://git.allucanget.biz/zwitschi/linearvoid/settings/actions/secrets`: `CI_REGISTRY_USER`, `CI_REGISTRY_TOKEN` (scope `write:packages`), `ADMIN_USERNAME`, `ADMIN_PASSWORD`, `DATABASE_URL`.
+- **Hetzner Cloud (optional, prod)** — US-East VPS (Ashburn, VA per ADR-001).
 
 ## Network Environments
 
-Linear Void operates in two distinct environments. Know which one you're targeting.
+Two distinct targets; know which one you're using.
 
 ```mermaid
 flowchart LR
     subgraph Test_Lab["Test Lab (192.168.88.0/24)"]
         TEST_HOST["192.168.88.40<br/>Docker Engine :2375"]
     end
-
     subgraph Prod["Production (Hetzner — NOT provisioned)"]
         PROD_HOST["Hetzner CX22 Ashburn, VA<br/>Public IP TBD"]
     end
-
     subgraph Gitea["Gitea CI/CD"]
         CI["CI: build + push images"]
         CD["CD: deploy via DOCKER_HOST"]
     end
-
     subgraph CF["Cloudflare"]
         DNS["linearvoid.tv"]
         R2["R2 Bucket"]
     end
-
     CD -->|"DOCKER_HOST=tcp://192.168.88.40:2375"| TEST_HOST
-    CD -.->|"DOCKER_HOST=tcp://&lt;hetzner-ip&gt;:2375<br/>(when provisioned)"| PROD_HOST
+    CD -.->|"DOCKER_HOST=tcp://&lt;hetzner-ip&gt;:2375 (when provisioned)"| PROD_HOST
     TEST_HOST -->|"test traffic"| DNS
     PROD_HOST -.->|"production traffic"| DNS
     TEST_HOST --> R2
     PROD_HOST -.-> R2
 ```
 
-**Test Lab (active):** Internal network `192.168.88.0/24`, host `192.168.88.40`. Receives automated CD deploys from Gitea Actions. Used for integration testing and staging validation.
-
-**Production (pending):** Hetzner Cloud VPS, Ashburn VA. Not yet provisioned. CD workflow and compose files are ready — only `DOCKER_HOST` needs changing. See Production Deployment section in [DEPLOYMENT.md](./DEPLOYMENT.md) for cutover plan.
-
-## Development Prerequisites
-
-- **Frontend dependencies**: Run `cd frontend && npm install` before `npm run dev` or `npm run build`.
-- **Docker build support**: Local development expects Docker to be able to build the `chat`, `playout`, and `nginx` images referenced in `docker-compose.yml`.
-- **Data directories**: Create `data/playlist`, `data/media`, `data/hls`, and `data/chat` before the first run so bind mounts resolve cleanly.
-- **Database choice**: The stack defaults to PostgreSQL in `docker-compose.yml`. If you want SQLite for local-only development, update `DATABASE_URL` in `.env` accordingly.
-- **Redis availability**: Chat presence, rate limiting, and pub/sub require Redis. The compose stack provides it automatically.
-- **Video inputs**: Source videos should be normalized before playout. If the playlist contains invalid files, the playout container will fail its validation flow.
+- **Test Lab (active):** internal `192.168.88.0/24`, host `192.168.88.40`; receives automated CD deploys from Gitea Actions. Used for staging validation.
+- **Production (pending):** Hetzner CX22, Ashburn VA — not yet provisioned. Compose/CD are ready; only `DOCKER_HOST` needs changing. See [DEPLOYMENT.md](./DEPLOYMENT.md).
 
 ## Environment Setup
 
-1. **Clone the Repository** (if not already done):
-
-   ```bash
-   git clone https://git.allucanget.biz/zwitschi/linearvoid.git
-   cd linearvoid
-   ```
-
-2. **Copy Environment File**:
-
-   ```bash
-   cp .env.example .env
-   ```
-
-3. **Configure `.env`** (copy from `.env.example` and replace placeholders)
-
-4. **Check local value overrides for Docker Compose**:
-   - `docker-compose.yml` builds images locally and maps the frontend build output from `frontend/dist` into nginx.
-   - `docker-compose.prod.yml` pulls prebuilt images from the Gitea registry.
-   - Internal container-to-container URLs should use compose hostnames, for example `REDIS_URL=redis://redis:6379/0` and `DATABASE_URL=postgres://linearvoid:<password>@db:5432/linearvoid?sslmode=disable`.
-   - For production public routing, set `VITE_PUBLIC_APP_URL=https://linearvoid.tv`, `VITE_PUBLIC_WS_URL=wss://linearvoid.tv/ws`, and `VITE_PUBLIC_HLS_URL=https://media.linearvoid.tv/hls/playlist.m3u8`.
-
-5. **Create Data Directories**:
-
-   ```bash
-   mkdir -p data/playlist data/media data/hls data/chat
-   ```
-
-6. **Prepare Sample Videos** (optional):
-   - Place normalized MP4 files in `data/media/`.
-   - Create a `data/playlist/playlist.txt` with paths (see ADR-001 for normalization requirements).
+1. `git clone https://git.allucanget.biz/zwitschi/linearvoid.git && cd linearvoid`
+2. `cp .env.example .env` then replace placeholders.
+3. Compose overrides: `docker-compose.yml` builds images locally and mounts `frontend/dist` into nginx; `docker-compose.prod.yml` pulls prebuilt images from Gitea.
+   - Internal URLs use compose hostnames: `REDIS_URL=redis://redis:6379/0`, `DATABASE_URL=postgres://linearvoid:<pw>@db:5432/linearvoid?sslmode=disable`.
+   - Public routing: `VITE_PUBLIC_APP_URL=https://linearvoid.tv`, `VITE_PUBLIC_WS_URL=wss://linearvoid.tv/ws`, `VITE_PUBLIC_HLS_URL=https://media.linearvoid.tv/hls/playlist.m3u8`.
+4. Create data dirs: `mkdir -p data/playlist data/media data/hls data/chat`
+5. Optional: place normalized MP4s in `data/media/` and create `data/playlist/playlist.txt` (see ADR-001 for normalization).
 
 ## Gitea Setup
 
-1. **Create or verify a Gitea user**:
-   - The user should have access to the repository and permission to read from the container registry.
-
-2. **Create an access token**:
-   - Generate a token with the scopes needed for package/registry access and, if applicable, repository automation.
-   - Store it as `CI_REGISTRY_TOKEN` in `.env`.
-
-3. **Configure Gitea Actions Secrets**:
-   - Go to the repository settings page: [https://git.allucanget.biz/zwitschi/linearvoid/settings/actions/secrets](https://git.allucanget.biz/zwitschi/linearvoid/settings/actions/secrets)
-   - Add the following secrets to enable automated CI/CD and production deployment:
-     - `CI_REGISTRY_USER`: Your Gitea username (e.g., `zwitschi`).
-     - `CI_REGISTRY_TOKEN`: Your Gitea Personal Access Token with `write:packages` scope.
-     - `ADMIN_USERNAME`: The administrator username for the production admin panel (e.g., `admin`).
-     - `ADMIN_PASSWORD`: A strong password for the production admin panel.
-     - `DATABASE_URL`: The production database connection string (e.g., `postgres://linearvoid:<password>@db:5432/linearvoid?sslmode=disable`).
-
-4. **Configure registry variables**:
-   - Set `CI_REGISTRY_URL` to the Gitea host.
-   - Set `CI_REGISTRY_NAMESPACE` to the image namespace, for example `zwitschi/linearvoid`.
-   - Set `CI_REGISTRY_USER` to the account that will push or pull images.
-
-5. **Authenticate Docker to the registry**:
-
-   ```bash
-   docker login git.allucanget.biz
-   ```
-
-6. **Build and publish images when needed**:
-   - Local development uses `docker-compose.yml`, which builds from source.
-   - Production uses `docker-compose.prod.yml`, which expects the `playout`, `chat`, and `nginx` images to exist in the configured Gitea registry namespace.
-
-7. **Validate image availability**:
-   - Before a production rollout, verify that the expected `:latest` tags can be pulled from the registry on the target server.
+1. User with repo + registry access; generate a token (scope `write:packages`) stored as `CI_REGISTRY_TOKEN`.
+2. Set Actions secrets (see Prerequisites) to enable automated CI/CD and deploys.
+3. Set registry vars: `CI_REGISTRY_URL`, `CI_REGISTRY_NAMESPACE` (e.g. `zwitschi/linearvoid`), `CI_REGISTRY_USER`.
+4. Auth: `docker login git.allucanget.biz`.
+5. Local dev builds from source; prod pulls `playout`, `chat`, `nginx` images from the registry namespace. Verify `:latest` tags exist before rollout.
 
 ## Cloudflare R2 Setup
 
-1. **Create an R2 bucket**:
-   - Create a bucket such as `linearvoid-media` in Cloudflare R2.
-   - Keep the bucket name aligned with `R2_BUCKET_NAME` in `.env`.
-
-2. **Generate R2 API credentials**:
-   - Create an access key pair with read/write permissions for the bucket.
-   - Set `R2_ACCESS_KEY_ID`, `R2_SECRET_ACCESS_KEY`, and `R2_REGION=auto` in `.env`.
-
-3. **Find the S3 endpoint**:
-   - Use the account-scoped endpoint format `https://<account-id>.r2.cloudflarestorage.com`.
-   - Store the account ID in both `R2_ENDPOINT` and `CLOUDFLARE_ACCOUNT_ID` where applicable.
-
-4. **Bind a custom media domain**:
-   - Attach a domain such as `media.linearvoid.tv` to the bucket.
-   - Set `MEDIA_DOMAIN` to that hostname.
-
-5. **Configure app-domain DNS and TLS**:
-   - Point `linearvoid.tv` to the public entrypoint serving nginx.
-   - Point `www.linearvoid.tv` to the same target and redirect it to `https://linearvoid.tv`.
-   - Keep `media.linearvoid.tv` on the R2 custom domain path so HLS delivery stays split from the app origin.
-   - Use Cloudflare proxied DNS with Full (strict) TLS so browser traffic uses `https://linearvoid.tv`, `wss://linearvoid.tv/ws`, and `https://media.linearvoid.tv`.
-
-6. **Apply cache behavior**:
-   - Configure the `.m4s` and `.m3u8` cache rules from [cache-rules](../docs/infrastructure/cloudflare/cache-rules.md).
-   - Enable Smart Tiered Cache for the R2-backed hostname.
-   - If you need header normalization for immutable segment caching, deploy `infrastructure/cloudflare/worker.js` to the same zone.
-
-7. **Configure R2 CORS**:
-   - Add CORS rules to allow the app domain to access the R2 bucket for HLS delivery.
-   - Example CORS configuration:
-
-   ```json
-   [
-     {
-       "AllowedOrigins": ["https://linearvoid.tv"],
-       "AllowedMethods": ["GET"],
-       "AllowedHeaders": ["*"],
-       "ExposeHeaders": ["ETag", "Cache-Control"],
-       "MaxAgeSeconds": 3600
-     }
-   ]
-   ```
-
-8. **Verify access**:
-   - Confirm uploads with an S3-compatible client such as `aws s3 ls --endpoint-url <R2_ENDPOINT>`.
-   - After deployment, check Cloudflare analytics for cache hit ratio and origin request volume.
-
-## Cloudflare
-
-Use this section for full edge setup (DNS, Pages, cache, security, and monitoring).
-
-Execution tracker: `docs/operations/cloudflare-uptimerobot-checklist.md`.
-
-### 1. DNS Setup
-
-1. Add apex record for `linearvoid.tv` pointing to Cloudflare Pages target.
-2. Add `www` CNAME to Cloudflare Pages and configure redirect to `https://linearvoid.tv`.
-3. Add `media.linearvoid.tv` for split-domain HLS delivery.
-4. Keep all records proxied (orange cloud) unless troubleshooting requires DNS-only mode.
-
-### 2. R2 Bucket
-
-1. Create bucket `linearvoid-media`.
-2. Set location hint to `enam` (US-East).
-3. Attach custom domain `media.linearvoid.tv` to the bucket.
-4. Configure CORS rules to allow `https://linearvoid.tv` for HLS delivery.
-5. Verify object read works from the custom domain.
-
-### 3. Cache Rules
-
-Configure Cloudflare Cache Rules:
-
-1. HLS segment files (`*.ts`, `*.m4s`):
-   - Cache: eligible
-   - Edge TTL: `31536000`
-   - Browser TTL: long-lived
-   - Cache key: include full path
-2. HLS manifests (`*.m3u8`):
-   - Cache: eligible
-   - Edge TTL: `2s`
-   - Revalidation: enabled
-3. Static assets (`*.js`, `*.css`):
-   - Cache: eligible
-   - Edge TTL: `1y`
-   - Browser TTL: long-lived
-   - Cache key: include full path
-4. Validate behavior using response headers:
-   - `Cf-Cache-Status` should move from `MISS` to `HIT` for hot objects
-
-Reference: [cache-rules](../docs/infrastructure/cloudflare/cache-rules.md).
-
-### 4. Cloudflare Pages
-
-Also see [the automated Gitea Actions deployment guide](./operations/gitea-cloudflare-pages.md).
-
-Manual setup:
-
-1. Connect Gitea repository to Cloudflare Pages.
-2. Build command:
-
-   ```bash
-   cd frontend && npm run build
-   ```
-
-3. Output directory:
-
-   ```text
-   frontend/dist
-   ```
-
-4. Configure production domain to `linearvoid.tv`.
-
-### 5. Tiered Cache
-
-1. Enable Smart Tiered Cache in Cloudflare dashboard.
-2. Warm a segment and manifest path.
-3. Confirm tiered behavior through repeated requests and `Cf-Cache-Status: HIT`.
-
-### 6. WAF and Bot Protection
-
-1. Enable OWASP Managed Ruleset.
-2. Start in `Simulate` mode.
-3. Review events for 48 hours.
-4. Switch to `Block` once false positives are addressed.
-5. Enable Bot Fight Mode.
-
-### 7. Rate Limiting
-
-Create rules:
-
-1. Path: `/api/admin/login`
-2. Path: `/api/admin/*`
-3. Threshold: `10 requests / 10 seconds / IP`
-4. Action: `Managed Challenge`
-
-### 8. HTTPS and WSS
-
-1. Enable `Always Use HTTPS`.
-2. Enable HSTS with:
-   - `max-age=31536000`
-   - `includeSubDomains`
-   - `preload`
-3. Minimum TLS version: `1.2`.
-4. Keep SSL mode at `Full (strict)` when origin certs are valid.
-5. Enforce HTTPS redirect at both layers:
-   - Cloudflare edge
-   - nginx origin
-6. Validate websocket endpoint uses `wss://linearvoid.tv/ws`.
-
-### 9. UptimeRobot Monitors
-
-Create external monitors:
-
-1. HTTP monitor:
-   - URL: `https://linearvoid.tv/health`
-   - Interval: 5 minutes
-2. Keyword monitor:
-   - URL: `https://linearvoid.tv/hls/playlist.m3u8`
-   - Keyword: `#EXTM3U`
-   - Interval: 5 minutes
-3. Alert policy:
-   - Trigger on 2 consecutive failures
-   - Notify via email and Discord/Slack webhook
-
-### 10. Verification Checklist
-
-1. DNS resolves for `linearvoid.tv`, `www.linearvoid.tv`, and `media.linearvoid.tv`.
-2. Pages serves frontend for apex and `www` redirects correctly.
-3. HLS segments are cacheable long-term, manifests refresh every ~2s.
-4. Admin paths are rate-limited and WAF events are visible.
-5. HTTPS and WSS both succeed with valid certificates.
-6. UptimeRobot alerts fire as expected when endpoints are intentionally made unavailable.
+1. Create R2 bucket (e.g. `linearvoid-media`, matching `R2_BUCKET_NAME`); generate an access key pair → `R2_ACCESS_KEY_ID`, `R2_SECRET_ACCESS_KEY`, `R2_REGION=auto`.
+2. S3 endpoint: `https://<account-id>.r2.cloudflarestorage.com` → `R2_ENDPOINT`.
+3. Attach `media.linearvoid.tv` as custom media domain → `MEDIA_DOMAIN`; add CORS allowing `https://linearvoid.tv` GET for HLS delivery.
+4. Point `linearvoid.tv` (+ `www` redirect) at the app entrypoint; keep `media.*` on R2. Proxied DNS, TLS **Full (strict)**.
+5. Apply the `.m4s`/`.m3u8` cache rules from [`cache-rules.md`](./infrastructure/cloudflare/cache-rules.md); enable Smart Tiered Cache; deploy the [`linearvoid-media` worker](../infrastructure/cloudflare/) on `media.linearvoid.tv` (replace the R2 bucket's direct custom-domain binding on that host) to enforce immutable segment caching + HLS CORS.
+6. Verify: `aws s3 ls --endpoint-url $R2_ENDPOINT`; check cache hit ratio after serving traffic.
 
 ## Running the Application
 
-### Local Development
+### Local development
 
-1. **Start All Services**:
+```bash
+docker compose up --build      # start all services
+docker compose down            # stop
+```
 
-   ```bash
-   docker compose up --build
-   ```
+Access: app/admin `https://linearvoid.tv` (via proxy) or `http://localhost`; WS `wss://linearvoid.tv/ws` / `ws://localhost/ws`. Health: playout `:8081/health`, chat `:3000/health`. Monitoring: Prometheus `:9090`, Grafana `:3001` (default `admin/admin`).
 
-2. **Access the Application**:
-   - Preferred local access with the documented proxy setup: <https://linearvoid.tv>
-   - Admin Interface: <https://linearvoid.tv/admin> (redirects to `/admin/login/` until authenticated)
-   - Chat WebSocket: `wss://linearvoid.tv/ws`
-   - Direct-local fallback when bypassing Cloudflare/Nginx Proxy Manager: <http://localhost>, <http://localhost/admin>, and `ws://localhost/ws`
-   - Health Checks:
-     - Playout: <http://localhost:8081/health>
-     - Chat: <http://localhost:3000/health>
-   - Monitoring:
-     - Prometheus: <http://localhost:9090>
-     - Grafana: <http://localhost:3001> (default login: admin/admin)
+Frontend dev (optional): `cd frontend && npm install && npm run dev` (http://localhost:5173, proxies API). For nginx-backed local runs: `npm run build` (nginx serves `frontend/dist`).
 
-3. **Frontend Development** (optional, for UI changes):
+### Test lab (remote Docker host)
 
-   ```bash
-   cd frontend
-   npm install
-   npm run dev
-   ```
+Deploy to `192.168.88.40` over Docker's TCP socket (port `2375`).
 
-   The dev server runs on <http://localhost:5173>. It proxies API calls to the backend services.
+Enable TCP on Alpine host — edit `/etc/conf.d/docker`:
 
-4. **Build frontend assets for nginx-backed local runs**:
+```bash
+DOCKER_OPTS="-H unix:///var/run/docker.sock -H tcp://0.0.0.0:2375"
+```
 
-   ```bash
-   cd frontend
-   npm install
-   npm run build
-   ```
+Then `rc-service docker restart`; verify with `netstat -tlnp | grep 2375`; allow port `2375` from `192.168.88.0/24` in the firewall.
 
-   The local nginx container serves `frontend/dist`, so rebuilding the frontend is required before validating production-like static output.
+Deploy from a machine with the Docker CLI:
 
-5. **Stop Services**:
+```bash
+export DOCKER_HOST="tcp://192.168.88.40:2375"
+docker info                                   # confirm remote host
+docker compose -f docker-compose.prod.yml pull
+docker compose -f docker-compose.prod.yml up -d --remove-orphans   # registry images
+# or build from source: docker compose up -d --build
+unset DOCKER_HOST                             # back to local
+```
 
-   ```bash
-   docker compose down
-   ```
+Health: `http://192.168.88.40:8081/health` (playout), `:3000/health` (chat), nginx entrypoint `:8080`.
 
-### Local Test Environment Deployment (Remote Docker Host)
+### Coolify deployment
 
-For testing deployments on a local network server without setting up a full production VPS, you can deploy directly to a remote Docker host (e.g., a home lab server or test machine) using Docker's TCP socket.
+Coolify replaces the Gitea CD `DOCKER_HOST` orchestration (Cloudflare DNS/R2/cache unchanged). Install (`curl -fsSL https://cdn.coollabs.io/coolify/install.sh | bash`), add the target server via SSH, create a Docker Compose resource from `docker-compose.prod.yml` (or the repo), configure env vars in Coolify's store, map persistent `data/` and named volumes, and assign `linearvoid.tv` to nginx for Traefik TLS. Full walkthrough: [`COOLIFY.md`](./COOLIFY.md).
 
-#### Test Environment Prerequisites
+### Production setup
 
-- A remote server running Docker at IP `192.168.88.40` (running Alpine Linux).
-- Docker TCP socket enabled and listening on port `2375` on the remote server.
+1. Provision a Linux VPS (Docker, Compose, Git); open HTTP/HTTPS ports.
+2. `git clone … && cd linearvoid && cp .env.example .env`.
+3. Set production `.env`: strong `POSTGRES_PASSWORD`/`GRAFANA_PASSWORD`, compose-hostname internal URLs, `MEDIA_DOMAIN=media.linearvoid.tv`, public `VITE_PUBLIC_*` URLs. For local-only testing bypassing the proxy, use `http://localhost`/`ws://localhost/ws`/`http://localhost/hls/playlist.m3u8`.
+4. `mkdir -p data/playlist data/media data/hls data/chat`.
+5. Ensure `playout`, `chat`, `nginx` images exist in the registry; `docker login` from the VPS.
+6. Deploy: `docker compose -f docker-compose.prod.yml pull && docker compose -f docker-compose.prod.yml up -d`.
+7. Verify: `docker compose -f docker-compose.prod.yml ps`, health endpoints, HLS written to `data/hls`, media domain serves via Cloudflare.
+8. Harden: HTTPS in front of nginx (Cloudflare, `www` → apex), restrict Grafana/Prometheus to trusted IPs, run under systemd for auto-start.
 
-##### Enabling Docker TCP Socket on Alpine Linux
+### Production (Hetzner VPS + Cloudflare)
 
-To allow the Gitea Actions runner (or your local machine) to deploy to the remote Alpine host, you must configure the Docker daemon to listen on the TCP socket:
+1. Provision CX22 in Ashburn, VA (ADR-001).
+2. Cloudflare: R2 bucket (`enam` hint), deploy the [`linearvoid-media` worker](../infrastructure/cloudflare/) on `media.linearvoid.tv` + cache rules, `linearvoid.tv`/`www` at the app entrypoint with `www`→apex redirect, Full (strict) TLS.
+3. Deploy from the project: `docker compose -f docker-compose.prod.yml up -d` (or pull from Gitea registry); manage under systemd.
+4. Monitoring: Grafana dashboards from `monitoring/grafana/`.
 
-1. Edit `/etc/conf.d/docker` on the Alpine host:
-
-   ```bash
-   vi /etc/conf.d/docker
-   ```
-
-2. Update the `DOCKER_OPTS` line to include both the unix socket and the TCP socket:
-
-   ```bash
-   DOCKER_OPTS="-H unix:///var/run/docker.sock -H tcp://0.0.0.0:2375"
-   ```
-
-3. Restart the Docker service:
-
-   ```bash
-   rc-service docker restart
-   ```
-
-4. Verify that Docker is listening on port `2375`:
-
-   ```bash
-   netstat -tlnp | grep 2375
-   ```
-
-5. Ensure any firewall (like `awall` or `iptables`) on the Alpine host allows incoming traffic on port `2375` from your local network subnet (e.g., `192.168.88.0/24`).
-
-#### Deployment Steps
-
-1. **Configure Docker CLI to use the Remote Host**:
-   You can point your local Docker CLI to the remote server by setting the `DOCKER_HOST` environment variable:
-   - **PowerShell (Windows)**:
-
-     ```powershell
-     $env:DOCKER_HOST="tcp://192.168.88.40:2375"
-     ```
-
-   - **Bash (Linux/macOS)**:
-
-     ```bash
-     export DOCKER_HOST="tcp://192.168.88.40:2375"
-     ```
-
-2. **Verify Connection**:
-   Run `docker info` to confirm your local CLI is communicating with the remote server at `192.168.88.40`:
-
-   ```bash
-   docker info
-   ```
-
-   The output should show the remote server's hostname and system details.
-
-3. **Deploy the Stack**:
-   With `DOCKER_HOST` set, any `docker` or `docker compose` commands you run locally will execute directly on the remote server.
-   - **Deploy from Registry (Production-like)**:
-
-     ```bash
-     docker compose -f docker-compose.prod.yml pull
-     docker compose -f docker-compose.prod.yml up -d --remove-orphans
-     ```
-
-   - **Build and Deploy from Source**:
-
-     ```bash
-     docker compose up -d --build
-     ```
-
-4. **Access the Test Environment**:
-   - Playout Health: `http://192.168.88.40:8081/health`
-   - Chat Health: `http://192.168.88.40:3000/health`
-   - Nginx Entrypoint: `http://192.168.88.40:8080` (or the port mapped to `NGINX_PORT` in `.env`)
-
-5. **Reset to Local Docker Host**:
-   Once testing is complete, clear the `DOCKER_HOST` variable to point back to your local machine:
-   - **PowerShell**:
-
-     ```powershell
-     Remove-Item env:DOCKER_HOST
-     ```
-
-   - **Bash**:
-
-     ```bash
-     unset DOCKER_HOST
-     ```
-
-### Production Setup Guide
-
-1. **Prepare the target host**:
-   - Provision a Linux VPS with Docker Engine, Docker Compose, Git, and enough disk for media, HLS output, PostgreSQL, Redis persistence, and monitoring data.
-   - Open the ports required for HTTP/HTTPS and any restricted monitoring access you intend to expose.
-
-2. **Clone the repository and create the environment file**:
-
-   ```bash
-   git clone https://git.allucanget.biz/zwitschi/linearvoid.git
-   cd linearvoid
-   cp .env.example .env
-   ```
-
-3. **Set production-safe `.env` values**:
-   - Replace placeholder secrets.
-   - Use compose-network hostnames for internal services, for example `redis` and `db`.
-   - Set strong values for `POSTGRES_PASSWORD` and `GRAFANA_PASSWORD`.
-   - Keep `MEDIA_DOMAIN`, R2 credentials, and Gitea registry values consistent with the external infrastructure.
-   - Set `MEDIA_DOMAIN=media.linearvoid.tv`.
-   - Set `VITE_PUBLIC_APP_URL=https://linearvoid.tv`, `VITE_PUBLIC_WS_URL=wss://linearvoid.tv/ws`, and `VITE_PUBLIC_HLS_URL=https://media.linearvoid.tv/hls/playlist.m3u8`.
-   - If you intentionally bypass the public proxy for local-only testing, switch those values to `http://localhost`, `ws://localhost/ws`, and `http://localhost/hls/playlist.m3u8`.
-
-4. **Create runtime directories on the server**:
-
-   ```bash
-   mkdir -p data/playlist data/media data/hls data/chat
-   ```
-
-5. **Publish or verify required images in Gitea**:
-   - Ensure the `playout`, `chat`, and `nginx` images exist under `${CI_REGISTRY_URL}/${CI_REGISTRY_NAMESPACE}`.
-   - Log in to the registry from the VPS before pulling.
-
-6. **Deploy the production stack**:
-
-   ```bash
-   docker compose -f docker-compose.prod.yml pull
-   docker compose -f docker-compose.prod.yml up -d
-   ```
-
-7. **Run post-deploy verification**:
-   - Check `docker compose -f docker-compose.prod.yml ps`.
-   - Confirm nginx, chat, and playout health endpoints respond.
-   - Verify that HLS output is being written to `data/hls` and that the media domain serves manifests and segments through Cloudflare.
-
-8. **Harden and operate the deployment**:
-   - Put HTTPS in front of nginx using Cloudflare or another reverse proxy, with `linearvoid.tv` as canonical app host and `www.linearvoid.tv` redirected.
-   - Restrict Grafana and Prometheus exposure to trusted IPs or a private network.
-   - Run the stack under systemd or another service manager if you need auto-start behavior outside Docker restart policies.
-
-### Production Deployment (Hetzner VPS + Cloudflare)
-
-1. **Provision VPS**: Use Hetzner Cloud CX22 in Ashburn, VA (per ADR-001).
-2. **Configure Cloudflare**:
-   - Create R2 bucket with `enam` location hint.
-   - Set up custom media domain (`media.linearvoid.tv`) with the cache rules from [cache-rules](../docs/infrastructure/cloudflare/cache-rules.md).
-   - Point `linearvoid.tv` and `www.linearvoid.tv` at the app entrypoint, then redirect `www` to the apex domain.
-   - Use Full (strict) TLS for both app and media hostnames.
-   - Apply the Worker from `infrastructure/cloudflare/worker.js` for advanced header management.
-3. **Deploy**:
-   - Copy the project to the VPS.
-   - Use production compose: `docker compose -f docker-compose.prod.yml up -d`
-   - Or pull from Gitea registry: `docker compose -f docker-compose.prod.yml pull && docker compose -f docker-compose.prod.yml up -d`
-   - Configure systemd to manage the Docker Compose stack.
-4. **Monitoring**: Access Grafana at the exposed port for dashboards (included in `monitoring/grafana/dashboard.json`).
-
-See `docs/architecture/README.md` for the full arc42 documentation and the ADRs for architectural decisions.
+See `docs/architecture/README.md` (arc42) and the ADRs for decisions and cost/scaling guidance (ADR-001).
 
 ## Troubleshooting
 
-- **R2 Upload Fails**: Check credentials in `.env` and ensure the bucket exists. Verify with `aws s3 ls --endpoint-url=$R2_ENDPOINT`.
-- **Normalization Errors**: Run `backend/playout/start-playout.sh` manually to see ffprobe output. All videos must be H.264/AAC, 30 fps constant, 2s keyframes (ADR-001).
-- **Chat Not Connecting**: Check Redis connection and WebSocket URL. The health check at `/health` should return 200.
-- **Cache Not Working**: Verify Cloudflare cache rules (`.m4s` = long TTL immutable, `.m3u8` = short TTL). Check cache hit ratio in Cloudflare dashboard.
-- **Docker Issues**: Run `docker compose logs` for service-specific errors. Ensure ports are not in use.
-- **Admin Interface**: Ensure `ADMIN_USERNAME`, `ADMIN_PASSWORD`, and `ADMIN_SESSION_TTL_HOURS` are set. `/admin/` and `/api/admin/*` are protected by a session-based login flow.
-- **Production Pull Fails**: Confirm `docker login` to Gitea succeeded on the server and the image tags referenced by `docker-compose.prod.yml` exist.
+- **R2 upload fails** — check `.env` creds/bucket; `aws s3 ls --endpoint-url=$R2_ENDPOINT`.
+- **Normalization errors** — run `backend/playout/start-playout.sh` to see ffprobe output. Video must be H.264/AAC, 30 fps constant, 2s keyframes (ADR-001).
+- **Chat won't connect** — check Redis and WS URL; `/health` should return 200.
+- **Cache not working** — verify `.m4s` long-TTL / `.m3u8` short-TTL cache rules and hit ratio.
+- **Docker issues** — `docker compose logs`; check port conflicts.
+- **Admin interface** — `ADMIN_USERNAME`, `ADMIN_PASSWORD`, `ADMIN_SESSION_TTL_HOURS` set; session-based login protects `/admin/` and `/api/admin/*`.
+- **Prod pull fails** — `docker login` to Gitea succeeded and image tags exist.
 
 ## Further Reading
 
-- Architecture: `docs/architecture/README.md` (arc42 template with diagrams).
-- Decisions: `docs/architecture/adr/` (ADR-001 to ADR-007).
-- Contributing: `CONTRIBUTING.md`.
-- Project Structure: Root `README.md`.
-
-The system is ready for use. For production, review ADR-001 for cost estimates and scaling guidance.
+Architecture: `docs/architecture/README.md` · Decisions: `docs/architecture/adr/` (ADR-001…007) · Contributing: `CONTRIBUTING.md` · Overview: root `README.md`.

docs/STREAMING.md

@@ -1,6 +1,6 @@
 # Streaming Pipeline
 
-This document is the operational contract for Linear Void video ingest, playout, publication, caching, and browser playback. General deployment, chat, and admin details remain in `DEPLOYMENT.md`, `SETUP.md`, and `API.md`.
+Operational contract for Linear Void video ingest, playout, publication, caching, and browser playback. Deployment, chat, admin details live in `DEPLOYMENT.md`, `SETUP.md`, `API.md`.
 
 ## Production flow
 
@@ -20,22 +20,22 @@ The playout service is the timeline authority. It reads the ordered playlist, ru
 https://media.linearvoid.tv/hls/playlist.m3u8
 ```
 
-Nginx `/hls/` is a local/test fallback, not the production media origin. The SPA, admin API, and chat use `linearvoid.tv`; media uses `media.linearvoid.tv`.
+Nginx `/hls/` is a local/test fallback, not the production origin. The SPA, admin API, and chat use `linearvoid.tv`; media uses `media.linearvoid.tv`.
 
 ## Source-media contract
 
 Only validated files may enter the production playlist.
 
-| Property   | Required profile                                                         |
-| ---------- | ------------------------------------------------------------------------ |
-| Container  | MP4                                                                      |
-| Video      | H.264/AVC, `yuv420p`                                                     |
-| Resolution | 1280x720 or 1920x1080; avoid changing resolution within a playlist block |
-| Frame rate | Constant 30 fps                                                          |
-| GOP        | Closed, fixed 2 seconds: `-g 60 -keyint_min 60 -sc_threshold 0`          |
-| Audio      | AAC-LC, stereo, 48 kHz, 128–192 kbit/s                                   |
+| Property   | Required profile                                                        |
+| ---------- | ----------------------------------------------------------------------- |
+| Container  | MP4                                                                     |
+| Video      | H.264/AVC, `yuv420p`                                                    |
+| Resolution | 1280x720 or 1920x1080; avoid resolution changes within a playlist block |
+| Frame rate | Constant 30 fps                                                         |
+| GOP        | Closed, fixed 2 seconds: `-g 60 -keyint_min 60 -sc_threshold 0`         |
+| Audio      | AAC-LC, stereo, 48 kHz, 128–192 kbit/s                                  |
 
-The ingest gate uses `ffprobe` to reject incompatible codecs, variable frame rate, missing audio, invalid dimensions, or incorrect keyframe cadence. Assets are immutable after acceptance; corrections create a new asset ID and object key.
+The ingest gate uses `ffprobe` to reject incompatible codecs, variable frame rate, missing audio, invalid dimensions, or wrong keyframe cadence. Assets are immutable after acceptance; corrections create a new asset ID/object key.
 
 Reference normalization:
 
@@ -51,140 +51,139 @@ ffmpeg -i input.mov \
 
 ## Playlist and timeline
 
-- `data/playlist/playlist.txt` contains one local media path per line; blank lines and `#` comments are ignored.
-- The service validates the complete playlist before activating a revision.
-- Playlist edits are written atomically and receive a revision ID. The active FFmpeg run keeps its current item and adopts the new revision at the next item boundary.
-- All viewers follow the same live manifest; clients do not calculate the schedule independently.
-- Schedule metadata must be derived from the same active playlist revision and timeline so “now/next” agrees with the video.
-- When the last item ends, playout returns to the first item unless an active schedule override says otherwise.
+- `data/playlist/playlist.txt`: one local media path per line; blank lines and `#` comments ignored.
+- Service validates the complete playlist before activating a revision.
+- Playlist edits are atomic and get a revision ID. The active FFmpeg run keeps its current item and adopts the new revision at the next item boundary.
+- All viewers follow the same live manifest; clients never calculate the schedule independently.
+- Schedule metadata derives from the same active revision/timeline so "now/next" agrees with the video.
+- When the last item ends, playout returns to the first unless an active schedule override says otherwise.
 
 ## Playout and HLS packaging
 
-The playout service wraps FFmpeg, tracks the active item and playlist revision, restarts failed processes, and exposes `/health` plus Prometheus metrics.
+The playout service wraps FFmpeg, tracks the active item/revision, restarts failed processes, and exposes `/health` plus Prometheus metrics.
 
-Baseline packaging:
+| Setting          | Value                                                                    |
+| ---------------- | ------------------------------------------------------------------------ |
+| Protocol         | Standard live HLS, not LL-HLS                                            |
+| Segment format   | fMP4 (`.m4s`) with init segment; `.ts` may be supported during migration |
+| Segment target   | 2 seconds (`HLS_SEGMENT_DURATION=2`)                                     |
+| Live window      | 12 segments (`HLS_WINDOW_SIZE=12`), ~24 seconds                          |
+| Codec path       | Stream copy for conforming assets; no per-viewer transcoding             |
+| Expected latency | 15–35 seconds                                                            |
 
-| Setting          | Value                                                                                 |
-| ---------------- | ------------------------------------------------------------------------------------- |
-| Protocol         | Standard live HLS, not LL-HLS                                                         |
-| Segment format   | fMP4 (`.m4s`) with an initialization segment; `.ts` may be supported during migration |
-| Segment target   | 2 seconds (`HLS_SEGMENT_DURATION=2`)                                                  |
-| Live window      | 12 segments (`HLS_WINDOW_SIZE=12`), about 24 seconds                                  |
-| Codec path       | Stream copy for conforming assets; no per-viewer transcoding                          |
-| Expected latency | 15–35 seconds                                                                         |
+Segment boundaries must align with keyframes; timestamps must stay monotonic across item boundaries. If FFmpeg cannot preserve a seamless timeline, insert `#EXT-X-DISCONTINUITY` and publish the corresponding new init segment.
 
-Segment boundaries must align with keyframes. Timestamps must remain monotonic across item boundaries. If FFmpeg cannot preserve a seamless timeline, insert `#EXT-X-DISCONTINUITY` and publish the corresponding new initialization segment.
+**Segment duration rationale**: 2s lowers latency and keeps the window small (12 × 2s = 24s) while staying above the manifest TTL (1–2s). The 2s GOP (30fps, `-g 60`) already matches the target, so segments cut cleanly on keyframes. Longer (4–6s) raises latency and window depth; shorter than 2s inflates segment/request overhead and R2 Class B ops with little gain.
 
-**Segment duration rationale**: 2s segments were chosen to lower end-to-end latency and keep the live window small (12 × 2s = 24s) while staying well above the manifest TTL (1–2s). The 2s GOP (30fps, `-g 60 -keyint_min 60`) already matches the target, so segments cut cleanly on keyframes. Longer durations (4–6s) increase latency and window depth; shorter than 2s inflates segment/request overhead and R2 Class B operations with little latency gain.
-
-Every segment name is globally unique and never reused, including after a restart or playlist loop. Use a run/session ID plus a monotonically increasing media sequence, for example:
+Every segment name is globally unique and never reused, including after restart or playlist loop — use a run/session ID plus a monotonically increasing media sequence:
 
 ```text
 hls/r-20260807T170412Z/init-000001.mp4
 hls/r-20260807T170412Z/seg-000001234.m4s
 ```
 
-This naming rule is essential: a URL advertised as immutable must never later contain different bytes.
+A URL advertised as immutable must never later contain different bytes.
 
 ## Safe publication to R2
 
-Publication order prevents a manifest from referencing an unavailable segment:
+Order prevents a manifest referencing an unavailable segment:
 
-1. FFmpeg writes a segment to a temporary local name and closes it.
-2. The publisher uploads the initialization segment, if new, and media segment to their final immutable R2 keys.
-3. It verifies successful upload and records object size/checksum.
-4. It renders the next media manifest referencing only confirmed objects.
-5. It uploads the manifest to a temporary key, then replaces `hls/playlist.m3u8` as the final commit step.
+1. FFmpeg writes a segment to a temp local name and closes it.
+2. Publisher uploads the init segment (if new) and media segment to final immutable R2 keys.
+3. Verify upload success; record object size/checksum.
+4. Render the next manifest referencing only confirmed objects.
+5. Upload the manifest to a temp key, then replace `hls/playlist.m3u8` as the final commit step.
 
-Never upload an actively written segment or publish the manifest first. Use a single publisher instance, or fencing/leader election, so two playout workers cannot update the live manifest concurrently.
+Never upload an actively written segment or publish the manifest first. Use a single publisher (or fencing/leader election) so two workers can't update the live manifest concurrently.
 
-The local `data/hls/` directory is staging data and may be regenerated. Source media, playlist revisions, and schedule state are authoritative; generated HLS is not backed up as application state.
+**Run-scoped manifest URIs.** Segments and the init file are stored run-scoped under `hls/r-{runID}/…`, so the live `hls/playlist.m3u8` must reference them with that run prefix. Before the final commit step, the publisher rewrites each bare relative media URI (segments plus `#EXT-X-MAP:URI`) to `r-{runID}/<name>`. `publishManifest` in `backend/playout/src/playlist.go` does this via `rewriteManifestToRun`, using the same `runID` under which the segments were uploaded; the rewrite is skipped in slate mode (its manifest is served as-is). The published manifest therefore contains, e.g.:
+
+```m3u8
+#EXT-X-MAP:URI="r-20260807T170412Z/init.mp4"
+#EXTINF:2.000000,
+r-20260807T170412Z/seg-000001234.m4s
+```
+
+The `linearvoid-media` Worker on `media.linearvoid.tv` is a faithful path→R2 mapper and serves these run-scoped paths directly; it does **not** rewrite URIs at serve time.
+
+`data/hls/` is regenerable staging. Source media, playlist revisions, and schedule state are authoritative; generated HLS is not backed up as application state.
 
 ## Cache contract
 
-| Objects                    | Edge policy                        | Browser policy                          | Required headers/behavior                                                    |
-| -------------------------- | ---------------------------------- | --------------------------------------- | ---------------------------------------------------------------------------- |
-| `.m4s`, `.ts`, init `.mp4` | Cache eligible; TTL 1 year         | TTL 1 year                              | `Cache-Control: public, max-age=31536000, immutable`; full path in cache key |
-| Live `.m3u8`               | Cache eligible; edge TTL 2 seconds | Do not retain beyond freshness interval | `Cache-Control: public, max-age=1, s-maxage=2, must-revalidate`              |
-| Versioned JS/CSS           | Cache eligible; TTL 1 year         | TTL 1 year                              | Content-hashed filenames and `immutable`                                     |
+| Objects                    | Edge policy                 | Browser policy                          | Required headers/behavior                                                    |
+| -------------------------- | --------------------------- | --------------------------------------- | ---------------------------------------------------------------------------- |
+| `.m4s`, `.ts`, init `.mp4` | Cache eligible; TTL 1 year  | TTL 1 year                              | `Cache-Control: public, max-age=31536000, immutable`; full path in cache key |
+| Live `.m3u8`               | Cache eligible; edge TTL 2s | Do not retain beyond freshness interval | `Cache-Control: public, max-age=1, s-maxage=2, must-revalidate`              |
+| Versioned JS/CSS           | Cache eligible; TTL 1 year  | TTL 1 year                              | Content-hashed filenames and `immutable`                                     |
 
 Further rules:
 
-- Do not add query-string cache busters to segment URLs; uniqueness belongs in the path.
-- Cache keys must preserve hostname and full path. Ignore only explicitly approved tracking parameters.
-- HLS responses require `Access-Control-Allow-Origin: https://linearvoid.tv` (or the reviewed origin allowlist), `GET`, `HEAD`, and `OPTIONS` support, and correct MIME types.
-- Enable Smart Tiered Cache for `media.linearvoid.tv` to reduce R2 reads on regional misses.
-- Purge manifests when necessary, but never purge immutable segments during normal operation.
-- Retain expired-window segments in R2 for at least the maximum manifest TTL plus client retry/grace time. A practical minimum is one hour; lifecycle deletion may run later according to storage policy.
+- No query-string cache busters on segment URLs; uniqueness belongs in the path.
+- Cache keys preserve hostname and full path; ignore only explicitly approved tracking params.
+- HLS responses need `Access-Control-Allow-Origin: https://linearvoid.tv` (or reviewed allowlist), `GET`/`HEAD`/`OPTIONS`, and correct MIME types.
+- Enable Smart Tiered Cache for `media.linearvoid.tv` to cut R2 reads on regional misses.
+- Purge manifests when necessary, never immutable segments during normal operation.
+- Retain expired-window segments at least max manifest TTL + retry/grace time; practical minimum 1 hour; lifecycle deletion may run later per storage policy.
 
-Expected validation:
+Validation:
 
 ```bash
 curl -I https://media.linearvoid.tv/hls/playlist.m3u8
 curl -I https://media.linearvoid.tv/hls/<segment>.m4s
 ```
 
-The manifest must refresh within a few seconds. Repeated segment requests should progress from `CF-Cache-Status: MISS` to `HIT`, retain the same ETag/content length, and carry the immutable cache header.
+The manifest must refresh within seconds; repeated segment requests should progress `CF-Cache-Status: MISS` → `HIT`, keep the same ETag/content length, and carry the immutable cache header.
 
 ## Player behavior
 
-The SPA uses native HLS where supported and hls.js elsewhere.
-
-- Firefox has no native HLS, so it uses the hls.js/MSE path; fMP4 (`.m4s` + init segment) is supported. Autoplay requires `muted` + `playsinline`, which are already set on the `<video>` element.
+SPA uses native HLS where supported, hls.js elsewhere.
 
+- Firefox has no native HLS → hls.js/MSE path; fMP4 (`.m4s` + init) is supported. Autoplay needs `muted` + `playsinline`, already set on the `<video>` element.
 - Start at the live edge with a conservative buffer; do not seek to playlist start.
-- Treat the manifest as the synchronization source and target roughly three segments behind the newest segment unless testing supports a smaller margin.
-- Recover transient network/media errors with bounded retries and exponential backoff.
-- On a fatal hls.js network error, reload the manifest; on a fatal media error, attempt media recovery once before recreating the player.
-- When tab suspension or drift leaves playback materially behind the target live position, seek forward to the target rather than playing the backlog.
-- Surface a clear reconnecting state. Never silently loop stale media.
-- Schedule/chat timestamps should account for the measured player latency rather than assuming wall-clock playout is already visible.
+- Treat the manifest as the sync source; target ~3 segments behind the newest unless testing supports a smaller margin.
+- Recover transient network/media errors with bounded retries + exponential backoff.
+- On fatal hls.js network error, reload the manifest; on fatal media error, attempt media recovery once before recreating the player.
+- If tab suspension/drift leaves playback materially behind the target, seek forward rather than play the backlog.
+- Surface a clear reconnecting state; never silently loop stale media.
+- Schedule/chat timestamps should account for measured player latency, not wall-clock playout.
 
 ## Failure and recovery
 
-The service retries an item failure with a bounded backoff. After the configured failure threshold it switches to a pre-normalized looping slate and reports `mode=slate`. It periodically retries normal playout and returns only after a segment and committed manifest have been published successfully.
+The service retries an item failure with bounded backoff. Past the failure threshold it switches to a pre-normalized looping slate and reports `mode=slate`, periodically retrying normal playout, returning only after a segment and committed manifest publish successfully.
 
-On process or host restart:
+On process/host restart:
 
-- create a new run/session ID and continue with globally unique segment paths;
-- resume from the schedule-derived current position when reliable state exists, otherwise start the current playlist item;
+- create a new run/session ID and continue with unique segment paths;
+- resume from schedule-derived current position when reliable state exists, else start the current playlist item;
 - never overwrite old segment objects;
-- keep the last valid manifest available until the replacement stream is playable;
-- emit a discontinuity when timeline continuity cannot be guaranteed.
+- keep the last valid manifest available until the replacement is playable;
+- emit a discontinuity when timeline continuity isn't guaranteed.
 
-Health is not merely “process running.” `/health` should report at least mode (`live`, `slate`, `starting`, `degraded`), active item, playlist revision, run ID, last segment sequence/time, last successful R2 publication, consecutive failures, and uptime. Mark the service unhealthy if the manifest or media sequence has not advanced for more than two segment durations.
+`/health` must report more than "process running": at least mode (`live`/`slate`/`starting`/`degraded`), active item, playlist revision, run ID, last segment sequence/time, last successful R2 publication, consecutive failures, uptime. Mark unhealthy if the manifest or media sequence hasn't advanced for more than two segment durations.
 
-## Monitoring and acceptance checks
+## Monitoring and acceptance
 
-Monitor and alert on:
+Monitor/alert on: manifest age and media-sequence advancement; time since last published segment; FFmpeg restarts and consecutive failures; live/slate mode; R2 upload latency/errors; segment 404/5xx and manifest/segment latency; Cloudflare cache-hit ratio, origin requests, R2 Class B ops; player startup, fatal errors, rebuffer ratio, live-edge drift; local staging disk usage.
 
-- manifest age and media-sequence advancement;
-- time since last successfully published segment;
-- FFmpeg restarts and consecutive item failures;
-- live/slate mode;
-- R2 upload latency and errors;
-- segment 404/5xx rate and manifest/segment request latency;
-- Cloudflare cache-hit ratio, origin requests, and R2 Class B operations;
-- player startup time, fatal errors, rebuffer ratio, and live-edge drift;
-- local staging disk usage.
-
-A release is streaming-ready only when one full playlist boundary has played without timestamp errors, the manifest never references a missing object, a second viewer joins the same timeline, cache headers match the table above, a forced FFmpeg failure activates the slate, and normal playout recovers without reusing an object URL.
+A release is streaming-ready only when: one full playlist boundary plays without timestamp errors; the manifest never references a missing object; a second viewer joins the same timeline; cache headers match the table; a forced FFmpeg failure activates slate; and normal playout recovers without reusing an object URL.
 
 ## Required configuration
 
-| Variable                                   | Purpose                                            |
-| ------------------------------------------ | -------------------------------------------------- |
-| `PLAYLIST_PATH`                            | Active playlist file                               |
-| `HLS_SEGMENT_DURATION`                     | Target segment duration; baseline `2`              |
-| `HLS_WINDOW_SIZE`                          | Manifest window; baseline `12`                     |
-| `R2_ENDPOINT`, `R2_REGION`                 | S3-compatible publication endpoint                 |
-| `R2_BUCKET_NAME`                           | Media bucket                                       |
-| `R2_ACCESS_KEY_ID`, `R2_SECRET_ACCESS_KEY` | Write credentials, available only to the publisher |
-| `MEDIA_DOMAIN`                             | Public R2 custom domain                            |
-| `VITE_PUBLIC_HLS_URL`                      | Public live manifest URL                           |
+| Variable                                   | Purpose                               |
+| ------------------------------------------ | ------------------------------------- |
+| `PLAYLIST_PATH`                            | Active playlist file                  |
+| `HLS_SEGMENT_DURATION`                     | Target segment duration; baseline `2` |
+| `HLS_WINDOW_SIZE`                          | Manifest window; baseline `12`        |
+| `R2_ENDPOINT`, `R2_REGION`                 | S3-compatible publication endpoint    |
+| `R2_BUCKET_NAME`                           | Media bucket                          |
+| `R2_ACCESS_KEY_ID`, `R2_SECRET_ACCESS_KEY` | Write credentials, publisher only     |
+| `R2_KEEP_RUNS`                             | Run folders to retain (default `2`)   |
+| `R2_MAX_RETENTION_BYTES`                   | Max bytes across run folders (10 GiB) |
+| `MEDIA_DOMAIN`                             | Public R2 custom domain               |
+| `VITE_PUBLIC_HLS_URL`                      | Public live manifest URL              |
 
-Production values are `MEDIA_DOMAIN=media.linearvoid.tv` and `VITE_PUBLIC_HLS_URL=https://media.linearvoid.tv/hls/playlist.m3u8`. Local fallback may use `http://localhost/hls/playlist.m3u8`.
+Production values: `MEDIA_DOMAIN=media.linearvoid.tv`, `VITE_PUBLIC_HLS_URL=https://media.linearvoid.tv/hls/playlist.m3u8`. Local fallback may use `http://localhost/hls/playlist.m3u8`.
 
 ## Explicitly deferred
 
-Adaptive-bitrate ladders, LL-HLS, DVR/catch-up, multi-origin playout, and regional chat affinity are outside the v1 contract. Adding any of them requires an ADR and an update to this document.
+ABR ladders, LL-HLS, DVR/catch-up, multi-origin playout, and regional chat affinity are outside the v1 contract. Adding any requires an ADR and an update to this document.

docs/architecture/adr/ADR-001.md

@@ -69,7 +69,7 @@ Adopt a **self-managed US-East playout + Cloudflare R2 + edge-cached HLS** archi
 
 ### 3. Frontend
 
-- Static SPA (HTML + custom CSS + vanilla JS) hosted on Cloudflare Pages or same VPS/Nginx.
+- Static SPA (HTML + custom CSS + vanilla JS) hosted on Cloudflare Workers or same VPS/Nginx.
 - Video player: `hls.js` (Chromium/Firefox) or native HLS (Safari). Single live `master.m3u8`.
 - Embedded chat sidebar with room switcher, private DMs, now-playing overlay.
 
@@ -82,7 +82,7 @@ Adopt a **self-managed US-East playout + Cloudflare R2 + edge-cached HLS** archi
 
 ### 5. Infrastructure Summary & Cost Model
 
-- **Primary**: Hetzner Ashburn VPS (playout + chat + Nginx) + Cloudflare R2 + Pages.
+- **Primary**: Hetzner Ashburn VPS (playout + chat + Nginx) + Cloudflare R2 + Workers.
 - **Monitoring**: MediaMTX/SRS dashboard, UptimeRobot, systemd logs, Cloudflare analytics.
 - **Backup**: Weekly R2/S3-compatible snapshots of playlist, DB, config.
 - **Fallback**: Emergency slate/loop playlist, automated restart.

docs/incident-runbook.md

@@ -4,155 +4,101 @@ Triage order for Linear Void incidents. Work top-to-bottom — each tier depends
 
 ## Triage Order
 
-1. Edge Reachability (Cloudflare)
-2. Origin Health (VPS/Docker)
-3. Playout Progression (HLS)
-4. Chat Connectivity (WebSocket)
-5. Data-Plane (Redis/DB)
+1. **Edge Reachability** (Cloudflare)
+2. **Origin Health** (VPS/Docker)
+3. **Playout Progression** (HLS)
+4. **Chat Connectivity** (WebSocket)
+5. **Data-Plane** (Redis/DB)
 
 ## 1. Edge Reachability (Cloudflare)
 
 **Symptoms:** `linearvoid.tv` not resolving, SSL errors, 5xx from Cloudflare.
 
 ```bash
-# Check DNS
 dig linearvoid.tv
 dig www.linearvoid.tv
-
-# Check Cloudflare status
 curl -I https://linearvoid.tv/health
 ```
 
-**Common causes:**
+**Common causes:** DNS misconfiguration (A record → wrong IP); Cloudflare proxy disabled (grey cloud exposes origin); SSL certificate expired (Cloudflare → SSL/TLS); Cloudflare outage ([cloudflarestatus.com](https://www.cloudflarestatus.com)).
 
-- DNS misconfiguration (A record pointing to wrong IP)
-- Cloudflare proxy disabled (grey cloud → origin exposed directly)
-- SSL certificate expired (check Cloudflare dashboard → SSL/TLS)
-- Cloudflare outage → check [https://www.cloudflarestatus.com](https://www.cloudflarestatus.com)
+**Mitigation:** Verify A record points `linearvoid.tv` → origin IP (proxied, orange cloud). Bypass Cloudflare temporarily:
 
-**Mitigation:**
-
-- Verify A record: `linearvoid.tv` → origin IP (proxied, orange cloud)
-- Bypass Cloudflare temporarily: `curl -H "Host: linearvoid.tv" http://<origin-ip>/health`
+```bash
+curl -H "Host: linearvoid.tv" http://<origin-ip>/health
+```
 
 ## 2. Origin Health (VPS/Docker)
 
-**Symptoms:** Cloudflare returns 502/504, containers down.
+**Symptoms:** Cloudflare 502/504, containers down.
 
 ```bash
-# Remote Docker (test lab)
 docker -H tcp://192.168.88.40:2375 ps
 docker -H tcp://192.168.88.40:2375 logs linearvoid-nginx --tail 50
-
-# Direct health check
 curl http://192.168.88.40/health
 ```
 
-**Common causes:**
-
-- Host out of disk space: `df -h`
-- Docker daemon stopped: `systemctl status docker`
-- Container restart loop: `docker ps -a --filter "status=exited"`
-- Resource exhaustion (OOM): `dmesg | grep -i oom`
+**Common causes:** Host disk full (`df -h`); Docker daemon stopped (`systemctl status docker`); container restart loop (`docker ps -a --filter "status=exited"`); OOM (`dmesg | grep -i oom`).
 
 **Mitigation:**
 
-- Restart stack: `docker compose -f docker-compose.prod.yml up -d`
-- Free disk: `docker system prune -f`
-- Check logs per service: `docker logs linearvoid-<service> --tail 100`
+```bash
+docker compose -f docker-compose.prod.yml up -d   # restart stack
+docker system prune -f                            # free disk
+docker logs linearvoid-<service> --tail 100       # per-service logs
+```
 
 ## 3. Playout Progression (HLS)
 
 **Symptoms:** Viewer sees slate or buffering, playlist not advancing.
 
 ```bash
-# Check playout health
-curl http://192.168.88.40:8081/health
-# Response includes "mode": "live" or "slate"
-
-# Check HLS output
+curl http://192.168.88.40:8081/health             # "mode": "live" or "slate"
 curl -I http://192.168.88.40/hls/playlist.m3u8
 ls -la data/hls/
-
-# Check playout logs
 docker logs linearvoid-playout --tail 100
 ```
 
-**Common causes:**
+**Common causes:** FFmpeg crash → slate fallback (health `degraded`); playlist empty/missing (`cat data/playlist/playlist.txt`); media file missing (`ls data/media/`); disk full → HLS segments can't write.
 
-- FFmpeg crash → slate fallback active (health shows `degraded`)
-- Playlist file empty/missing: `cat data/playlist/playlist.txt`
-- Media file missing from playlist: `ls data/media/`
-- Disk full → HLS segments can't write
-
-**Mitigation:**
-
-- Verify playlist has valid entries
-- Check media files exist at paths in playlist
-- Restart playout: `docker compose restart playout`
-- Generate demo playlist if empty: see `backend/playout/start-playout.sh`
+**Mitigation:** Validate playlist entries and media paths; restart playout (`docker compose restart playout`); regenerate demo playlist if empty (see `backend/playout/start-playout.sh`).
 
 ## 4. Chat Connectivity (WebSocket)
 
 **Symptoms:** Chat not loading, messages not sending, presence broken.
 
 ```bash
-# Check chat health
 curl http://192.168.88.40:3000/health
-
-# Test WebSocket
 wscat -c ws://192.168.88.40/ws?username=test&room=main
-
-# Check chat logs
 docker logs linearvoid-chat --tail 100
 ```
 
-**Common causes:**
+**Common causes:** Redis down → presence/rate-limiting degraded (chat still works); DB down → messages not persisting; port not exposed (`docker port linearvoid-chat`).
 
-- Redis down → presence/rate-limiting degraded (chat still works)
-- Database down → messages not persisting
-- Port not exposed: `docker port linearvoid-chat`
-
-**Mitigation:**
-
-- Restart chat: `docker compose restart chat`
-- Check Redis: `docker exec linearvoid-redis redis-cli ping`
-- Check DB: `docker exec linearvoid-db pg_isready -U linearvoid`
-- Chat works without Redis (graceful degradation) — check logs for "Redis connection warning"
+**Mitigation:** Restart chat (`docker compose restart chat`); check Redis (`docker exec linearvoid-redis redis-cli ping`); check DB (`docker exec linearvoid-db pg_isready -U linearvoid`). Chat degrades gracefully without Redis — look for "Redis connection warning" in logs.
 
 ## 5. Data-Plane (Redis/DB)
 
 **Symptoms:** Chat works but no persistence/presence, admin stats empty.
 
 ```bash
-# Redis
 docker exec linearvoid-redis redis-cli ping
 docker exec linearvoid-redis redis-cli info memory
-
-# PostgreSQL
 docker exec linearvoid-db pg_isready -U linearvoid
 docker exec linearvoid-db psql -U linearvoid -c "SELECT count(*) FROM messages;"
 ```
 
-**Common causes:**
+**Common causes:** Redis OOM (check `maxmemory`); PostgreSQL connection limit; disk full on volume mount.
 
-- Redis OOM (check `maxmemory` setting)
-- PostgreSQL connection limit reached
-- Disk full on volume mount
-
-**Mitigation:**
-
-- Redis: `docker compose restart redis` (AOF persist enabled)
-- PostgreSQL: `docker compose restart db`
-- Restore from backup: see `scripts/verify-backups.sh` for procedure
+**Mitigation:** Restart redis (`docker compose restart redis`, AOF persist enabled); restart db (`docker compose restart db`); restore from backup (see `scripts/verify-backups.sh`).
 
 ## Escalation
 
-| Severity | When                                 | Action                                             |
-| -------- | ------------------------------------ | -------------------------------------------------- |
-| Critical | Slate active >30min, complete outage | Trigger DR plan, provision fallback host           |
-| Warning  | Redis down, high CPU >30min          | Investigate during business hours, restart service |
-| Info     | Single service blip                  | Monitor, no action unless recurring                |
+| Severity | When                                 | Action                                      |
+| -------- | ------------------------------------ | ------------------------------------------- |
+| Critical | Slate active >30min, complete outage | Trigger DR plan, provision fallback host    |
+| Warning  | Redis down, high CPU >30min          | Investigate business hours, restart service |
+| Info     | Single service blip                  | Monitor; no action unless recurring         |
 
 ## Post-Incident
 
@@ -163,67 +109,52 @@ docker exec linearvoid-db psql -U linearvoid -c "SELECT count(*) FROM messages;"
 
 ## Quarterly Disaster Recovery Drill
 
-Goal: validate recovery targets under controlled host-loss simulation.
-
-**Recovery targets:** RPO <= 24h, RTO <= 60min.
+Validate recovery targets under controlled host-loss simulation: **RPO ≤ 24h, RTO ≤ 60min**.
 
 ### Prerequisites
 
-- Latest backup artifacts available from `scripts/backup-db.sh` and `scripts/backup-weekly.sh`
+- Latest backups from `scripts/backup-db.sh` and `scripts/backup-weekly.sh`
 - Operator access to Docker host and DNS controls
-- Current deployment variables available (`.env`, compose file, image tags)
+- Current deployment variables (`.env`, compose file, image tags)
 
-### Drill Scope
+### Scope
 
 - Simulate complete origin loss (all app containers unavailable)
 - Rebuild stack on clean host
 - Restore DB and media/playlist mounts
 - Verify viewer playback, chat, admin auth, monitoring targets
 
-### Drill Steps
+### Steps
 
 1. Record drill start timestamp and on-call participants.
 2. Capture pre-drill baseline:
-   - `curl -fsS http://<origin>/health`
-   - `curl -fsS http://<origin>:3000/health`
-   - `curl -fsS http://<origin>:8081/health`
-3. Simulate origin outage:
-   - `docker compose -f docker-compose.prod.yml down`
-4. Provision clean recovery host (or clean test-lab VM) and copy deployment bundle.
-5. Restore required data paths:
-   - `data/playlist/`
-   - `data/media/`
-   - `data/chat/`
-   - `data/hls/` (optional warm cache)
+   ```bash
+   curl -fsS http://<origin>/health
+   curl -fsS http://<origin>:3000/health
+   curl -fsS http://<origin>:8081/health
+   ```
+3. Simulate outage: `docker compose -f docker-compose.prod.yml down`
+4. Provision clean recovery host (or clean test-lab VM); copy deployment bundle.
+5. Restore data paths: `data/playlist/`, `data/media/`, `data/chat/`, `data/hls/` (optional warm cache).
 6. Restore database from latest backup dump.
-7. Start stack:
-   - `docker compose -f docker-compose.prod.yml up -d`
-8. Validate service health and runtime:
-   - `docker ps --format "table {{.Names}}\t{{.Status}}"`
-   - `curl -fsS http://<origin>/health`
-   - `curl -fsS http://<origin>/hls/playlist.m3u8 | head -n 5`
-   - `curl -fsS http://<origin>:9090/-/ready`
-9. Validate app behavior:
-   - Viewer SPA loads and HLS plays
-   - WebSocket chat connects and message round-trip works
-   - Admin login/session endpoint returns success
-10. Run backup integrity check:
-    - `scripts/verify-backups.sh`
-11. Record drill end timestamp and compute measured RTO/RPO.
+7. Start stack: `docker compose -f docker-compose.prod.yml up -d`
+8. Validate health and runtime:
+   ```bash
+   docker ps --format "table {{.Names}}\t{{.Status}}"
+   curl -fsS http://<origin>/health
+   curl -fsS http://<origin>/hls/playlist.m3u8 | head -n 5
+   curl -fsS http://<origin>:9090/-/ready
+   ```
+9. Validate app behavior: viewer SPA loads and HLS plays; WebSocket chat connects with message round-trip; admin login/session endpoint succeeds.
+10. Run backup integrity check: `scripts/verify-backups.sh`
+11. Record drill end timestamp; compute measured RTO/RPO.
 
 ### Pass/Fail Criteria
 
-- Pass if all core services healthy and functional in <= 60 minutes.
-- Pass if latest recoverable data point is <= 24 hours old.
-- Fail if any critical path remains degraded (playout, chat, admin auth, monitoring).
+- **Pass:** all core services healthy and functional in ≤ 60 min.
+- **Pass:** latest recoverable data point ≤ 24 h old.
+- **Fail:** any critical path remains degraded (playout, chat, admin auth, monitoring).
 
 ### Evidence Template
 
-Document each drill in incident tracker with:
-
-- Date/time window
-- Environment (test-lab, staging, production)
-- Backup artifact IDs and timestamps
-- Measured RTO and RPO
-- Issues found and remediation owner
-- Follow-up ticket links
+Record in incident tracker: date/time window; environment (test-lab, staging, production); backup artifact IDs and timestamps; measured RTO and RPO; issues found and remediation owner; follow-up ticket links.

docs/infrastructure/cloudflare/cloudflare-account-troubleshooting.md

@@ -2,23 +2,23 @@
 
 Reference for Cloudflare account and API token troubleshooting.
 
-Use `https://api.cloudflare.com/client/v4/accounts/${CLOUDFLARE_ACCOUNT_ID}/pages/projects` to verify the account ID and API token are valid.
+Use `wrangler whoami` to verify the account ID and API token are valid, or query the Workers API at `https://api.cloudflare.com/client/v4/accounts/${CLOUDFLARE_ACCOUNT_ID}/workers/scripts`.
 
 ## Validate Cloudflare account and API token
 
-```powershell
+```bash
 # Validate Cloudflare account and API token in .env
-Set-Location -Path "E:\Documents\02-Projects\07-Mentoring\Corey\linearvoid"
-Get-Content .env | Select-String 'CLOUDFLARE_API_TOKEN'
-Get-Content .env | Select-String 'CLOUDFLARE_ACCOUNT_ID'
+cd /path/to/linearvoid
+grep -E 'CLOUDFLARE_API_TOKEN|CLOUDFLARE_ACCOUNT_ID' .env
 
-# Test API token with curl
-$env:CLOUDFLARE_API_TOKEN = (Get-Content .env | Select-String 'CLOUDFLARE_API_TOKEN' | ForEach-Object { $_.Line.Split('=')[1].Trim() })
-$env:CLOUDFLARE_ACCOUNT_ID = (Get-Content .env | Select-String 'CLOUDFLARE_ACCOUNT_ID' | ForEach-Object { $_.Line.Split('=')[1].Trim() })
+# Test API token with wrangler
+npx wrangler whoami
 
-# List all Pages projects for the account
-curl -X GET "https://api.cloudflare.com/client/v4/accounts/$env:CLOUDFLARE_ACCOUNT_ID/pages/projects" -H "Authorization: Bearer $env:CLOUDFLARE_API_TOKEN" -H "Content-Type: application/json"
+# Or verify with the Workers API directly
+export CLOUDFLARE_API_TOKEN="$(grep '^CLOUDFLARE_API_TOKEN=' .env | cut -d= -f2-)"
+export CLOUDFLARE_ACCOUNT_ID="$(grep '^CLOUDFLARE_ACCOUNT_ID=' .env | cut -d= -f2-)"
 
-# Create a new Pages project (if needed)
-# curl -X POST "https://api.cloudflare.com/client/v4/accounts/$env:CLOUDFLARE_ACCOUNT_ID/pages/projects" -H "Authorization: Bearer $env:CLOUDFLARE_API_TOKEN" -H "Content-Type: application/json" -d '{"name":"linearvoid","production_branch":"main"}'
+# List all Workers for the account
+curl -X GET "https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/workers/scripts" \
+  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" -H "Content-Type: application/json"
 ```

docs/operations/cloudflare-pages-setup.md

@@ -1,69 +0,0 @@
-# Cloudflare Pages & Workers Setup — Linear Void
-
-Linear Void uses two Cloudflare products:
-
-| Product     | Purpose            | What                                                                                 |
-| ----------- | ------------------ | ------------------------------------------------------------------------------------ |
-| **Pages**   | Static SPA hosting | Vite-built frontend (`frontend/dist`) deployed via Git integration or Direct Upload  |
-| **Workers** | Edge cache logic   | `infrastructure/cloudflare/worker.js` — adds immutable cache headers to HLS segments |
-
-Both are managed under **Workers & Pages** in the Cloudflare dashboard.
-
-## Dashboard URLs
-
-| Resource                 | URL                                                                                                                 |
-| ------------------------ | ------------------------------------------------------------------------------------------------------------------- |
-| Workers & Pages overview | `https://dash.cloudflare.com/deb11491c54ec53a24772e98ac6d8128/workers-and-pages`                                    |
-| Worker: linearvoid       | `https://dash.cloudflare.com/deb11491c54ec53a24772e98ac6d8128/workers/services/view/linearvoid/production/settings` |
-| Pages project            | `https://dash.cloudflare.com/deb11491c54ec53a24772e98ac6d8128/pages/view/linearvoid`                                |
-| API tokens               | <https://dash.cloudflare.com/profile/api-tokens>                                                                    |
-
-## Worker: `linearvoid` (edge cache headers)
-
-The Worker at `infrastructure/cloudflare/worker.js` runs on the `linearvoid` Workers service. It intercepts requests and adds immutable `Cache-Control` headers to `.m4s` segment files.
-
-Deploy via Wrangler CLI:
-
-```bash
-npx wrangler deploy infrastructure/cloudflare/worker.js
-```
-
-Or paste into the Workers dashboard at the URL above (Quick Edit).
-
-## Pages: Static frontend
-
-### Direct Upload via Gitea Actions
-
-Use `.gitea/workflows/pages-deploy.yml` to build in Gitea Actions and upload via the Cloudflare API.
-
-See [gitea-cloudflare-pages.md](./gitea-cloudflare-pages.md) for the workflow.
-
-### Variables and secrets
-
-On Cloudflare pages, set the following environment variables (needed to run the frontend)
-
-| Type      | Name                                   | Value                                           |
-| --------- | -------------------------------------- | ----------------------------------------------- |
-| Secret    | ADMIN_PASSWORD                         | `not-a-real-admin-password`                     |
-| Plaintext | ADMIN_SESSION_TTL_HOURS                | `12`                                            |
-| Plaintext | ADMIN_USERNAME                         | `boogers`                                       |
-| Plaintext | HLS_SEGMENT_DURATION                   | `2`                                             |
-| Plaintext | HLS_WINDOW_SIZE                        | `12`                                            |
-| Plaintext | PLAYLIST_PATH                          | `/app/playlist/playlist.txt`                    |
-| Plaintext | VITE_PUBLIC_ADMIN_DEFAULT_MODERATOR    | `boogers`                                       |
-| Plaintext | VITE_PUBLIC_ADMIN_DEFAULT_ROOM         | `main`                                          |
-| Plaintext | VITE_PUBLIC_ADMIN_MODERATION_LOG_LIMIT | `50`                                            |
-| Plaintext | VITE_PUBLIC_APP_URL                    | `https://linearvoid.tv`                         |
-| Plaintext | VITE_PUBLIC_HLS_URL                    | `https://media.linearvoid.tv/hls/playlist.m3u8` |
-| Plaintext | VITE_PUBLIC_WS_URL                     | `wss://linearvoid.tv/ws`                        |
-
-## Custom domains
-
-- `linearvoid.tv` → Pages project
-- `www.linearvoid.tv` → redirect to `linearvoid.tv`
-- `media.linearvoid.tv` → R2 bucket (separate; not Pages)
-
-## Related
-
-- [Gitea to Cloudflare Pages deployment guide](./gitea-cloudflare-pages.md)
-- [Cloudflare & UptimeRobot checklist](./cloudflare-uptimerobot-checklist.md)

docs/operations/cloudflare-uptimerobot-checklist.md

@@ -5,7 +5,7 @@ Purpose: execute and record all manual dashboard tasks that cannot be automated
 Recommended execution order:
 
 1. Bind `media.linearvoid.tv` and confirm the HLS origin is separate from the app origin.
-2. Enable Pages, Tiered Cache, and cache behavior for the media zone.
+2. Enable Workers, Tiered Cache, and cache behavior for the media zone.
 3. Apply WAF, bot, rate limiting, HTTPS, and HSTS settings.
 4. Verify monitor targets and alert delivery in UptimeRobot.
 5. Capture evidence and fill in the sign-off block.
@@ -15,18 +15,18 @@ Recommended execution order:
 - [ ] Pending
 - [x] Done
 
-## 1) Cloudflare Pages Link and Build
+## 1) Cloudflare Workers Deploy and Custom Domain
 
-- [ ] Connect repo to Cloudflare Pages project
+- [ ] Create a Cloudflare Worker for the SPA and bind the build output
 - [ ] Build command set to: `cd frontend && npm run build`
 - [ ] Output directory set to: `frontend/dist`
 - [ ] Production domain attached to `linearvoid.tv`
 
 Evidence to capture:
 
-- Screenshot of Pages deployment settings
+- Screenshot of Worker deployment settings
 - Deployment URL and latest deployment timestamp
-- Confirmed production build output path in the Pages project
+- Confirmed production build output path in the Worker
 
 ## 2) Tiered Cache Enablement and Validation
 
@@ -125,7 +125,7 @@ Evidence to capture:
 
 Completion summary:
 
-- Pages build URL:
+- Worker build URL:
 - Cache validation result:
 - WAF/rate-limit status:
 - HTTPS/HSTS status:

docs/operations/cloudflare-workers-setup.md

@@ -0,0 +1,71 @@
+# Cloudflare Workers Setup — Linear Void
+
+Linear Void uses Cloudflare Workers for two purposes:
+
+| Product     | Purpose            | What                                                                                         |
+| ----------- | ------------------ | -------------------------------------------------------------------------------------------- |
+| **Workers** | Static SPA hosting | `linearvoid-frontend` (`frontend/wrangler.toml`) — Vite SPA (`frontend/dist`) static assets  |
+| **Workers** | Edge HLS delivery  | `linearvoid-media` (`infrastructure/cloudflare/`) — R2-bound Worker on `media.linearvoid.tv` |
+
+Both are managed under **Workers & Pages** in the Cloudflare dashboard.
+
+## Dashboard URLs
+
+| Resource                 | URL                                                                                                                       |
+| ------------------------ | ------------------------------------------------------------------------------------------------------------------------- |
+| Workers & Pages overview | `https://dash.cloudflare.com/deb11491c54ec53a24772e98ac6d8128/workers-and-pages`                                          |
+| Worker: linearvoid-media | `https://dash.cloudflare.com/deb11491c54ec53a24772e98ac6d8128/workers/services/view/linearvoid-media/production/settings` |
+| API tokens               | <https://dash.cloudflare.com/profile/api-tokens>                                                                          |
+
+## Worker: `linearvoid-media` (edge HLS delivery)
+
+The Worker in [`infrastructure/cloudflare/`](../../infrastructure/cloudflare/) runs on the `linearvoid-media` Workers service. It owns the `media.linearvoid.tv` host, serves HLS objects from the R2 bucket (bound as `MEDIA_BUCKET`), and stamps immutable `Cache-Control` on segments plus a short TTL on live manifests (CORS + Range supported). See its [README](../../infrastructure/cloudflare/README.md).
+
+**Run-scoped layout.** R2 objects are stored run-scoped (`hls/r-{runID}/seg-*.m4s`, `…/init.mp4`), and the live `hls/playlist.m3u8` references them with the run prefix (the playout rewrites bare URIs before publishing — see [`docs/STREAMING.md`](../STREAMING.md)). The Worker is a faithful path→R2 mapper: it serves whatever key the request path resolves to (e.g. `/hls/r-…/seg-*.m4s`), adds the cache headers by file type, and does not rewrite manifest URIs at serve time.
+
+Test / deploy:
+
+```bash
+cd infrastructure/cloudflare
+npm install && npm test              # unit + Miniflare R2 tests
+npx wrangler deploy                  # preview/test target
+npx wrangler deploy --env production # attach media.linearvoid.tv
+```
+
+Attach the production custom domain only after removing `media.linearvoid.tv` from the R2 bucket's direct custom-domain binding. Automated production deploy: [`.gitea/workflows/workers-media-deploy.yml`](../../.gitea/workflows/workers-media-deploy.yml).
+
+## Worker: static frontend
+
+Serve the Vite-built SPA (`frontend/dist`) as Worker static assets:
+
+1. Build the frontend: `cd frontend && npm run build`.
+2. Bind `frontend/dist` to the Worker as static assets (Workers Static Assets / `assets` in `wrangler.toml`).
+3. Attach the `linearvoid.tv` custom domain to the Worker.
+
+See [gitea-cloudflare-workers.md](./gitea-cloudflare-workers.md) for the automated deployment workflow.
+
+### Build-time variables
+
+The `VITE_PUBLIC_*` values are baked into the bundle at build time by Vite (`vite.config.ts` uses `envDir: ".."` to read them from the repo root). They are supplied by the Gitea Actions build environment from Gitea secrets — not as Worker runtime variables. The static-assets Worker has no runtime JavaScript, so it reads no `env` at request time.
+
+| Name                                     | Value                                           |
+| ---------------------------------------- | ----------------------------------------------- |
+| `VITE_PUBLIC_APP_URL`                    | `https://linearvoid.tv`                         |
+| `VITE_PUBLIC_WS_URL`                     | `wss://linearvoid.tv/ws`                        |
+| `VITE_PUBLIC_HLS_URL`                    | `https://media.linearvoid.tv/hls/playlist.m3u8` |
+| `VITE_PUBLIC_ADMIN_DEFAULT_MODERATOR`    | `boogers`                                       |
+| `VITE_PUBLIC_ADMIN_DEFAULT_ROOM`         | `main`                                          |
+| `VITE_PUBLIC_ADMIN_MODERATION_LOG_LIMIT` | `50`                                            |
+
+Backend services (playout, chat, nginx) read their own runtime variables (`ADMIN_*`, `HLS_*`, `PLAYLIST_PATH`, `R2_*`, `DATABASE_URL`, `REDIS_URL`) from the container environment — see [`.env.example`](../../.env.example).
+
+## Custom domains
+
+- `linearvoid.tv` → Worker (static SPA, `linearvoid-frontend`)
+- `www.linearvoid.tv` → redirect to `linearvoid.tv`
+- `media.linearvoid.tv` → Worker (`linearvoid-media`, R2-bound) serving the `linearvoid-media` bucket
+
+## Related
+
+- [Gitea to Cloudflare Workers deployment guide](./gitea-cloudflare-workers.md)
+- [Cloudflare & UptimeRobot checklist](./cloudflare-uptimerobot-checklist.md)

docs/operations/gitea-cloudflare-pages.md

@@ -1,144 +0,0 @@
-# Gitea to Cloudflare Pages — Linear Void
-
-## Goal
-
-Deploy the Linear Void frontend (Vite + TypeScript SPA) from Gitea to Cloudflare Pages using Gitea Actions.
-
-## Prerequisites
-
-> **Important:** The Pages project must exist before the API deploy call works. You'll get error 8000007 ("Project not found") if you skip this step.
-
-1. **Cloudflare Pages project** created first:
-   - Dashboard → Workers & Pages → Create → Pages → Direct Upload
-   - Name it `linearvoid` (or a different name if `linearvoid` is already taken by a Worker — see note below)
-   - See [Cloudflare Pages Setup](./cloudflare-pages-setup.md) for full details.
-2. **Project name**: note the exact name (e.g., `linearvoid`).
-3. **API token** with Pages permissions:
-   - Create at <https://dash.cloudflare.com/profile/api-tokens>
-   - Permission: `Account → Cloudflare Pages → Edit`
-   - Store in Gitea → Settings → Secrets as `CLOUDFLARE_API_TOKEN`
-4. **Account ID** from Cloudflare dashboard overview → store as `CLOUDFLARE_ACCOUNT_ID`
-
-> **Naming conflict:** If a Worker already exists with the name `linearvoid`, the Pages project must use a different name (e.g., `linearvoid-site`). Update the `projectName` in the workflow accordingly. Workers and Pages share a namespace under "Workers & Pages".
-
-## Project-specific values
-
-| Setting            | Value                                                      |
-| ------------------ | ---------------------------------------------------------- |
-| Framework          | Vite + TypeScript                                          |
-| Build command      | `cd frontend && npm install && npm run build`              |
-| Output directory   | `frontend/dist`                                            |
-| Node version       | 20                                                         |
-| Multi-page entries | `index.html`, `admin/index.html`, `admin/login/index.html` |
-
-## Gitea Actions workflow
-
-Create `.gitea/workflows/pages-deploy.yml`:
-
-```yaml
-name: Deploy to Cloudflare Pages
-
-on:
-  push:
-    branches: ["main"]
-  workflow_dispatch: {}
-
-jobs:
-  deploy:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-
-      - name: Setup Node
-        uses: actions/setup-node@v4
-        with:
-          node-version: 22
-
-      - name: Install dependencies
-        run: cd frontend && npm ci
-
-      - name: Build
-        run: cd frontend && npm run build
-
-      - name: Deploy to Cloudflare Pages
-        uses: cloudflare/pages-action@v1
-        with:
-          apiToken: ${{ secrets.CLOUDFLARE_API_TOKEN }}
-          accountId: ${{ secrets.CLOUDFLARE_ACCOUNT_ID }}
-          projectName: linearvoid
-          directory: frontend/dist
-          gitHubToken: ${{ secrets.GITEA_TOKEN }}
-```
-
-If the `cloudflare/pages-action@v1` action is unavailable from Gitea, use the direct Cloudflare Pages API (multipart/form-data with manifest):
-
-Snippet to generate a manifest in format `-F manifest='{"index.html": "abc123", "style.css": "def456"}'`
-
-```bash
-# Build manifest form argument from all files in frontend/dist
-form_args="-F manifest={"
-for file in $(find frontend/dist -type f | sed 's|^frontend/dist/||'); do
-  hash=$(sha256sum "frontend/dist/$file" | awk '{print $1}')
-  form_args="$form_args\"$file\":\"$hash\","
-done
-form_args="${form_args%,}}" # remove trailing comma and close JSON
-```
-
-```yaml
-- name: Generate deployment manifest
-  run: |
-    cd frontend/dist
-    form_args="-F manifest={"
-    for file in $(find . -type f ! -name manifest.json | sed 's|^\./||'); do
-      hash=$(sha256sum "$file" | awk '{print $1}')
-      form_args="$form_args\"$file\":\"$hash\","
-    done
-    form_args="${form_args%,}}" # remove trailing comma and close JSON
-    echo "Form arguments: $form_args"
-    echo "FORM_ARGS=$form_args" >> $GITHUB_ENV
-
-- name: Deploy to Cloudflare Pages
-  env:
-    CLOUDFLARE_API_TOKEN: ${{ secrets.CLOUDFLARE_API_TOKEN }}
-    CLOUDFLARE_ACCOUNT_ID: ${{ secrets.CLOUDFLARE_ACCOUNT_ID }}
-  run: |
-    cd frontend/dist
-    form_args="$FORM_ARGS" # use form arguments generated in previous step
-    COMMIT_HASH=$(git rev-parse --short HEAD 2>/dev/null || echo "unknown")
-    curl -sS -X POST \
-      "https://api.cloudflare.com/client/v4/accounts/${CLOUDFLARE_ACCOUNT_ID}/pages/projects/linearvoid/deployments" \
-      -H "Authorization: Bearer ${CLOUDFLARE_API_TOKEN}" \
-      -F "commit_hash=${COMMIT_HASH}" \
-      $form_args
-```
-
-This matches the Cloudflare Pages Direct Upload API: `Content-Type: multipart/form-data`, each file as a separate form field, and a `manifest` JSON mapping filenames to SHA-256 hashes.
-
-## Secrets to add in Gitea
-
-| Secret                  | Value                                |
-| ----------------------- | ------------------------------------ |
-| `CLOUDFLARE_API_TOKEN`  | API token with Pages:Edit permission |
-| `CLOUDFLARE_ACCOUNT_ID` | Cloudflare account ID from dashboard |
-
-## Verify deployment
-
-1. Push to `main` or trigger manually from Gitea Actions.
-2. Check **Gitea Actions** → job logs for the deploy step.
-3. Check **Cloudflare Pages** → Deployments for the new deployment.
-4. Visit the `.pages.dev` preview URL, then promote to `linearvoid.tv` when ready.
-
-## Notes
-
-- The existing `ci.yml` already builds the frontend; the Pages deploy workflow is additive and independent.
-- Multi-page SPA (`vite.config.ts`) produces three entry points — Cloudflare Pages serves them correctly.
-- For SPA fallback routing, add a `frontend/public/_redirects` file with `/* /index.html 200` before building.
-- If deployments fail with 403, verify the API token has `Account → Cloudflare Pages → Edit` permission.
-- If you get error **8000007** ("Project not found"), the Pages project hasn't been created yet. Create it via the Cloudflare dashboard (Workers & Pages → Create → Pages → Direct Upload) first, then re-run.
-- If `linearvoid` is already taken by a Worker, use a different project name like `linearvoid-site` and update the `projectName` and `PROJECT_NAME` values in the workflow.
-
-## Related resources
-
-- Cloudflare Workers dashboard: `https://dash.cloudflare.com/deb11491c54ec53a24772e98ac6d8128/workers/services/view/linearvoid/production/settings` — for managing the `infrastructure/cloudflare/worker.js` deployment
-- Cloudflare Pages project: `https://dash.cloudflare.com/deb11491c54ec53a24772e98ac6d8128/pages/view/linearvoid` — deployment history and custom domain settings

docs/operations/gitea-cloudflare-workers.md

@@ -0,0 +1,107 @@
+# Gitea to Cloudflare Workers — Linear Void
+
+## Goal
+
+Deploy the Linear Void frontend (Vite + TypeScript SPA) from Gitea to Cloudflare Workers using Gitea Actions and Wrangler.
+
+## Prerequisites
+
+1. **Cloudflare Worker** for the SPA:
+   - Dashboard → Workers & Pages → Create → Worker (or Workers Static Assets).
+   - Name it `linearvoid-frontend` (the edge-cache Worker already uses the `linearvoid` name).
+2. **API token** with Workers permissions:
+   - Create at <https://dash.cloudflare.com/profile/api-tokens>
+   - Permission: `Account → Workers Scripts → Edit` and `Account → Workers Routes → Edit`
+   - Store in Gitea → Settings → Secrets as `CLOUDFLARE_API_TOKEN`
+3. **Account ID** from Cloudflare dashboard overview → store as `CLOUDFLARE_ACCOUNT_ID`
+4. **Wrangler config** (`frontend/wrangler.toml`) binding the `frontend/dist` output as static assets.
+
+> **Naming:** Workers share a namespace under "Workers & Pages". The media edge Worker is `linearvoid-media` ([`infrastructure/cloudflare/`](../../infrastructure/cloudflare/wrangler.toml), deployed via `.gitea/workflows/workers-media-deploy.yml`) and the SPA Worker is `linearvoid-frontend` (`frontend/wrangler.toml`, deployed via `.gitea/workflows/workers-deploy.yml`).
+
+## Project-specific values
+
+| Setting            | Value                                                      |
+| ------------------ | ---------------------------------------------------------- |
+| Framework          | Vite + TypeScript                                          |
+| Build command      | `cd frontend && npm install && npm run build`              |
+| Output directory   | `frontend/dist`                                            |
+| Node version       | 22                                                         |
+| Multi-page entries | `index.html`, `admin/index.html`, `admin/login/index.html` |
+
+## Gitea Actions workflow
+
+Build the SPA and deploy with Wrangler:
+
+```yaml
+name: Deploy to Cloudflare Workers
+
+on:
+  workflow_run:
+    workflows: [CI]
+    types: [completed]
+  workflow_dispatch: {}
+
+jobs:
+  deploy:
+    if: ${{ github.event_name == 'workflow_dispatch' || github.event.workflow_run.conclusion == 'success' }}
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Setup Node
+        uses: actions/setup-node@v4
+        with:
+          node-version: 22
+
+      - name: Install dependencies
+        run: cd frontend && npm ci
+
+      - name: Build
+        env:
+          VITE_PUBLIC_APP_URL: ${{ secrets.VITE_PUBLIC_APP_URL }}
+          VITE_PUBLIC_WS_URL: ${{ secrets.VITE_PUBLIC_WS_URL }}
+          VITE_PUBLIC_HLS_URL: ${{ secrets.VITE_PUBLIC_HLS_URL }}
+          VITE_PUBLIC_ADMIN_DEFAULT_MODERATOR: ${{ secrets.VITE_PUBLIC_ADMIN_DEFAULT_MODERATOR }}
+          VITE_PUBLIC_ADMIN_DEFAULT_ROOM: ${{ secrets.VITE_PUBLIC_ADMIN_DEFAULT_ROOM }}
+          VITE_PUBLIC_ADMIN_MODERATION_LOG_LIMIT: "50"
+        run: cd frontend && npm run build
+
+      - name: Deploy to Cloudflare Workers
+        env:
+          CLOUDFLARE_API_TOKEN: ${{ secrets.CLOUDFLARE_API_TOKEN }}
+          CLOUDFLARE_ACCOUNT_ID: ${{ secrets.CLOUDFLARE_ACCOUNT_ID }}
+        run: |
+          cd frontend
+          npx wrangler deploy
+```
+
+## Secrets to add in Gitea
+
+| Secret                                | Value                                |
+| ------------------------------------- | ------------------------------------ |
+| `CLOUDFLARE_API_TOKEN`                | API token with Workers permissions   |
+| `CLOUDFLARE_ACCOUNT_ID`               | Cloudflare account ID from dashboard |
+| `VITE_PUBLIC_APP_URL`                 | Public viewer URL                    |
+| `VITE_PUBLIC_WS_URL`                  | WebSocket endpoint                   |
+| `VITE_PUBLIC_HLS_URL`                 | HLS playlist URL                     |
+| `VITE_PUBLIC_ADMIN_DEFAULT_MODERATOR` | Default moderator username           |
+| `VITE_PUBLIC_ADMIN_DEFAULT_ROOM`      | Default chat room                    |
+
+## Verify deployment
+
+1. Push to `main` or trigger manually from Gitea Actions.
+2. Check **Gitea Actions** → job logs for the deploy step.
+3. Check **Cloudflare Workers** → `linearvoid-frontend` → Deployments for the new version.
+4. Visit `https://linearvoid-frontend.zwitschi82.workers.dev`, then attach `linearvoid.tv` when ready.
+
+## Notes
+
+- The existing `ci.yml` already builds the frontend; the Workers deploy workflow is additive and independent.
+- Multi-page SPA (`vite.config.ts`) produces three entry points — Workers static assets serve them correctly.
+- For SPA fallback routing, set Workers static assets `not_found_handling` to `single-page-application` (serve `index.html` for unknown paths).
+- The deploy workflow lives in `.gitea/workflows/workers-deploy.yml` (triggered after CI succeeds).
+
+## Related resources
+
+- Cloudflare Workers dashboard: `https://dash.cloudflare.com/deb11491c54ec53a24772e98ac6d8128/workers/services/view/linearvoid/production/settings` — for managing the `infrastructure/cloudflare/worker.js` deployment and the SPA Worker

docs/operations/scaling-performance-playbook.md

@@ -1,345 +1,156 @@
 # Scaling and Performance Playbook
 
-Purpose: define repeatable method to measure performance, capture baseline data, and decide scaling actions.
-
-Scope: planning and execution guide for TODO items under "Scaling & Performance (planning only)".
+Purpose: define a repeatable method to measure performance, capture baselines, and decide scaling actions for the "Linear Void" streaming app. Scope: planning/execution for the "Scaling & Performance (planning only)" TODO items.
 
 ## 1) Manifest TTL Tuning Plan
 
-Goal: tune `.m3u8` TTL to about `0.5x` segment duration while keeping playback stable.
+Goal: tune the `.m3u8` TTL to ~`0.5x` segment duration while keeping playback stable.
 
-Inputs:
-
-- Segment duration (`HLS_SEGMENT_DURATION`)
-- Current manifest TTL (Cloudflare cache rule)
-- Cache outcome distribution (`Cf-Cache-Status`)
+Inputs: segment duration (`HLS_SEGMENT_DURATION`), current manifest TTL (Cloudflare cache rule), cache outcome distribution (`Cf-Cache-Status`).
 
 Procedure:
-
 1. Start from current manifest TTL (2s).
 2. Capture 24h baseline headers from repeated `playlist.m3u8` requests.
 3. Test candidate TTLs: 1s, 2s, 3s.
-4. Compare:
-   - rebuffer events from player logs
-   - origin request increase/decrease
-   - HIT ratio changes
-5. Select lowest TTL that keeps playback stable and origin load acceptable.
+4. Compare rebuffer events (player logs), origin request increase/decrease, and HIT ratio changes.
+5. Select the lowest TTL that keeps playback stable and origin load acceptable.
 
-Acceptance:
-
-- No material increase in rebuffer rate.
-- Manifest staleness not observable by users.
+Acceptance: no material increase in rebuffer rate; manifest staleness not observable by users.
 
 ## 2) TTFB Baseline Plan
 
 Goal: measure warm-cache TTFB from 3+ US regions with target `< 4s`.
 
-Suggested regions:
-
-- US East (Virginia)
-- US Central
-- US West (California/Oregon)
+Regions: US East (Virginia), US Central, US West (California/Oregon).
 
 Method:
-
 1. Use WebPageTest with repeat view enabled.
 2. Run at least 5 samples per region.
 3. Record median and p95 TTFB.
-4. Store results in table (date, region, median, p95, test URL).
+4. Store results in a table (date, region, median, p95, test URL).
 
-Acceptance:
-
-- Warm-cache median < 4s in all tested regions.
-- Outliers documented with likely cause.
+Acceptance: warm-cache median < 4s in all regions; outliers documented with likely cause.
 
 ### TTFB Baseline Worksheet
+Use one row per region; keep the raw WebPageTest result URL with the summary for comparison.
 
-Use one row per region and keep the raw WebPageTest result URL with the summary so later runs can be compared against the same evidence.
+- US East (Virginia) / US Central / US West (California/Oregon): Date/time, Test URL `https://linearvoid.tv`, WebPageTest result URL, Median TTFB, p95 TTFB, Notes.
 
-- US East (Virginia)
-  - Date/time:
-  - Test URL: <https://linearvoid.tv>
-  - WebPageTest result URL:
-  - Median TTFB:
-  - p95 TTFB:
-  - Notes:
-- US Central
-  - Date/time:
-  - Test URL: <https://linearvoid.tv>
-  - WebPageTest result URL:
-  - Median TTFB:
-  - p95 TTFB:
-  - Notes:
-- US West (California/Oregon)
-  - Date/time:
-  - Test URL: <https://linearvoid.tv>
-  - WebPageTest result URL:
-  - Median TTFB:
-  - p95 TTFB:
-  - Notes:
-
-Suggested capture notes:
-
-- Use repeat view for the warm-cache measurement.
-- Keep the same browser profile and test URL across regions.
-- Note any CDN cache-rule changes or deploys that occurred between samples.
+Capture notes: use repeat view for warm-cache; keep the same browser profile and test URL across regions; note any CDN cache-rule changes or deploys between samples.
 
 ## 3) WebSocket Load-Test Plan
 
-Goal: characterize chat service behavior across 10 -> 100 -> 500 -> 1000 concurrent clients.
+Goal: characterize chat behavior across 10 → 100 → 500 → 1000 concurrent clients.
 
 Tooling option A (Artillery):
-
 ```bash
 npx artillery run scripts/loadtest/ws-chat.yml
 ```
 
-Tooling option B (custom Go harness):
+Tooling option B (custom Go harness): open N concurrent WS clients, send a message every X seconds, measure round-trip latency and failures.
 
-- open N concurrent WebSocket clients
-- send message every X seconds
-- measure round-trip latency and failures
+Metrics: p50/p95 message latency; error rate (connect/send/receive); dropped connections; CPU/memory for `chat`, `redis`, `nginx`.
 
-Metrics to capture:
-
-- p50/p95 message latency
-- error rate (connect, send, receive)
-- dropped connections
-- CPU/memory for `chat`, `redis`, `nginx`
-
-Acceptance:
-
-- p95 chat latency <= 1s at target concurrency tier.
-- error rate <= 1% during steady state.
+Acceptance: p95 chat latency ≤ 1s at target concurrency tier; error rate ≤ 1% during steady state.
 
 ### WebSocket Load-Test Worksheet
-
-- Date/time:
-- Environment:
-- Commit SHA / image tag:
-- Test URL:
-- Tool used:
-- Scenario duration:
-- Concurrency tiers reached:
-- p50 message latency:
-- p95 message latency:
-- Error rate:
-- Dropped connections:
-- CPU/memory notes for chat:
-- CPU/memory notes for redis:
-- CPU/memory notes for nginx:
-- Raw results location:
-- Conclusion:
-- Next action:
+Date/time, Environment, Commit SHA/image tag, Test URL, Tool used, Scenario duration, Concurrency tiers reached, p50 message latency, p95 message latency, Error rate, Dropped connections, CPU/memory notes for chat/redis/nginx, Raw results location, Conclusion, Next action.
 
 ## 4) Admin API Benchmark Plan
 
 Goal: measure admin API p95 and identify slow endpoints.
 
-Tooling option A (`k6`):
+Endpoints: `/api/admin/overview`, `/api/admin/analytics`, `/api/admin/chat-log`, `/api/admin/users`.
 
+Tooling option A (`k6`):
 ```bash
 k6 run scripts/loadtest/admin-api.js
 ```
 
 Tooling option B (`wrk`):
-
 ```bash
 wrk -t4 -c100 -d60s http://<host>/api/admin/overview
 ```
 
-Endpoints:
-
-- `/api/admin/overview`
-- `/api/admin/analytics`
-- `/api/admin/chat-log`
-- `/api/admin/users`
-
-Acceptance:
-
-- p95 < 1s for primary admin endpoints.
-- slow endpoint list with remediation notes.
+Acceptance: p95 < 1s for primary admin endpoints; slow endpoint list with remediation notes.
 
 ### Admin API Benchmark Worksheet
-
-- Date/time:
-- Environment:
-- Commit SHA / image tag:
-- Test URL:
-- Tool used:
-- Scenario duration:
-- Concurrency / virtual users:
-- `/api/admin/overview` p95:
-- `/api/admin/analytics` p95:
-- `/api/admin/chat-log` p95:
-- `/api/admin/users` p95:
-- Error rate:
-- Slowest endpoint:
-- Cache or DB notes:
-- Raw results location:
-- Conclusion:
-- Next action:
+Date/time, Environment, Commit SHA/image tag, Test URL, Tool used, Scenario duration, Concurrency/virtual users, `/api/admin/overview|analytics|chat-log|users` p95, Error rate, Slowest endpoint, Cache or DB notes, Raw results location, Conclusion, Next action.
 
 ## 5) Cache Hit Ratio Tracking Plan
 
 Goal: track Cloudflare cache hit ratio and alert when below threshold.
 
-Implementation outline:
-
+Implementation:
 1. Pull Cloudflare GraphQL analytics on schedule.
-2. Export to Prometheus-compatible metric path.
-3. Add Grafana panel with 5m/1h/24h windows.
+2. Export to a Prometheus-compatible metric path.
+3. Add a Grafana panel with 5m/1h/24h windows.
 4. Alert when hit ratio < 90% for 15m.
 
 Reference exporter:
-
 - `scripts/cache-hit-ratio.js`
-- `scripts/cache-hit-ratio.ps1` for Windows scheduled tasks / manual runs
-- Example run: `CLOUDFLARE_API_TOKEN=... CLOUDFLARE_ZONE_TAG=... CLOUDFLARE_HOSTNAME=media.linearvoid.tv node scripts/cache-hit-ratio.js`
-- Windows example: `pwsh -File scripts/cache-hit-ratio.ps1`
+- `scripts/cache-hit-ratio.sh` for scheduled tasks / manual runs
+- Example run: `export CLOUDFLARE_API_TOKEN=... CLOUDFLARE_ZONE_TAG=... CLOUDFLARE_HOSTNAME=media.linearvoid.tv; node scripts/cache-hit-ratio.js`
 - Intended output: Prometheus text format on stdout, suitable for cron, a Node wrapper, or a small scrape endpoint.
 
-Metric target:
-
-- `cache_hit_ratio >= 0.90`
+Metric target: `cache_hit_ratio >= 0.90`.
 
 ### Cache Hit Ratio Worksheet
-
-- Date/time:
-- Environment:
-- Commit SHA / image tag:
-- Hostname:
-- Window minutes:
-- Cloudflare zone tag:
-- Export command:
-- Raw exporter output location:
-- Hit requests:
-- Total requests:
-- Cache hit ratio:
-- Threshold met:
-- Notes:
+Date/time, Environment, Commit SHA/image tag, Hostname, Window minutes, Cloudflare zone tag, Export command, Raw exporter output location, Hit requests, Total requests, Cache hit ratio, Threshold met, Notes.
 
 ## 6) Capacity Trigger Rules
 
-Define operational triggers:
-
+Operational triggers:
 - Cache miss ratio > 10% for 15m
 - Chat p95 latency > 1s for 5m
 - Host CPU > 80% for 10m
 - Redis memory usage ratio (`linearvoid_chat_redis_memory_usage_ratio`) > 0.85 for 10m
 
 Action mapping:
-
 - Level 1: investigate + restart non-critical components
 - Level 2: scale service resources (CPU/RAM)
 - Level 3: activate standby planning path
 
 ### Capacity Trigger Review
-
-- Date/time:
-- Environment:
-- Commit SHA / image tag:
-- Trigger hit:
-- Threshold value:
-- Observed value:
-- Duration above threshold:
-- Level selected:
-- Immediate action taken:
-- Services involved:
-- Follow-up owner:
-- Raw metrics / alert link:
-- Notes:
+Date/time, Environment, Commit SHA/image tag, Trigger hit, Threshold value, Observed value, Duration above threshold, Level selected, Immediate action taken, Services involved, Follow-up owner, Raw metrics/alert link, Notes.
 
 ## 7) Phase 2 Service Separation Plan
 
-Objective: separate bottlenecked services while keeping simple ops.
+Objective: separate bottlenecked services while keeping ops simple.
 
 Boundaries:
-
 - `playout` isolated first (CPU-heavy)
 - `chat` + `redis` grouped for latency
 - `db` isolated for IO stability
 - monitoring stack optionally separated after sustained load
 
-Initial sizing guidance:
-
+Initial sizing:
 - playout host: 4 vCPU / 8 GB
-- chat host: 2-4 vCPU / 4-8 GB
+- chat host: 2–4 vCPU / 4–8 GB
 - db host: 2 vCPU / 4 GB + fast disk
 
-Exit criteria for Phase 2:
-
-- repeated trigger events over 7 days
-- capacity saturation despite tuning
+Exit criteria: repeated trigger events over 7 days; capacity saturation despite tuning.
 
 ### Phase 2 Boundary Worksheet
-
-- Date/time:
-- Environment:
-- Commit SHA / image tag:
-- Trigger history summary:
-- Repeated trigger count over 7 days:
-- Primary bottleneck:
-- Proposed split:
-- `playout` target host size:
-- `chat` target host size:
-- `db` target host size:
-- Monitoring separation needed:
-- Data/state migration notes:
-- Implementation owner:
-- Review decision:
-- Next action:
+Date/time, Environment, Commit SHA/image tag, Trigger history summary, Repeated trigger count over 7 days, Primary bottleneck, Proposed split, `playout`/`chat`/`db` target host size, Monitoring separation needed, Data/state migration notes, Implementation owner, Review decision, Next action.
 
 ## 8) Phase 3 Warm Standby Plan
 
 Objective: second-region resilience with DNS failover.
 
-Candidate layout:
-
+Layout:
 - Primary: Hetzner Ashburn (US-East)
 - Standby: Hetzner Hillsboro (US-West)
 
-Design:
-
-- Warm standby stack deployed and health-checked
-- Redis replication or fallback mode documented
-- Postgres replication strategy selected (streaming replica or periodic snapshot restore)
-- Cloudflare DNS failover policy defined (health-based or manual emergency switch)
-
-Failover drill requirements:
+Design: warm standby stack deployed and health-checked; Redis replication or fallback mode documented; Postgres replication strategy selected (streaming replica or periodic snapshot restore); Cloudflare DNS failover policy defined (health-based or manual emergency switch).
 
 ### Phase 3 Warm Standby Worksheet
-
-- Date/time:
-- Environment:
-- Commit SHA / image tag:
-- Primary region:
-- Standby region:
-- DNS failover mode:
-- Redis strategy:
-- Postgres strategy:
-- Health check endpoints:
-- Simulated cutover performed:
-- Measured RTO:
-- Measured RPO:
-- Open risks:
-- Implementation owner:
-- Review decision:
-- Notes:
+Date/time, Environment, Commit SHA/image tag, Primary region, Standby region, DNS failover mode, Redis strategy, Postgres strategy, Health check endpoints, Simulated cutover performed, Measured RTO, Measured RPO, Open risks, Implementation owner, Review decision, Notes.
 
 ## 9) Data Capture Template
-
-For each benchmark run, capture:
-
-- Date/time
-- Environment (test-lab/staging/production)
-- Commit SHA / image tag
-- Test scenario and duration
-- Raw metrics (latency, errors, resource usage)
-- Conclusion and next action
+For each benchmark run capture: Date/time, Environment (test-lab/staging/production), Commit SHA/image tag, Test scenario and duration, Raw metrics (latency, errors, resource usage), Conclusion and next action.
 
 ## 10) Execution Order
-
 1. TTFB baseline
 2. WebSocket load profile
 3. Admin API benchmark
@@ -349,14 +160,9 @@ For each benchmark run, capture:
 7. Phase 3 standby design review
 
 ## 11) Benchmark Run Sheet
+Use the same worksheet for every baseline/re-test so results stay comparable.
 
-Use the same worksheet for every baseline or re-test so results stay comparable.
-
-Preconditions:
-
-- Confirm the target environment and commit SHA/image tag.
-- Confirm Cloudflare cache rules are active when testing TTFB or manifest TTL.
-- Confirm the chat, admin, and playout services are healthy before load testing.
+Preconditions: confirm target environment and commit SHA/image tag; confirm Cloudflare cache rules are active when testing TTFB or manifest TTL; confirm chat, admin, and playout services are healthy before load testing.
 
 Run template:
 
@@ -373,70 +179,47 @@ Run template:
 | Next action                |       |
 
 Suggested command mapping:
-
 - TTFB baseline: WebPageTest repeat-view run against `https://linearvoid.tv`.
 - WebSocket profile: `npx artillery run scripts/loadtest/ws-chat.yml` or the Go harness referenced above.
 - Admin API benchmark: `k6 run scripts/loadtest/admin-api.js` or `wrk -t4 -c100 -d60s http://<host>/api/admin/overview`.
 
 ## 12) Scaling Decision Gate
 
-Treat the following as the default review gate after any benchmark batch:
-
+Default review gate after any benchmark batch:
 1. If warm-cache TTFB meets target but cache hit ratio falls below 90%, tune cache behavior before increasing capacity.
 2. If chat or admin latency exceeds target while CPU and memory remain low, prioritize query or handler optimization.
-3. If CPU, memory, or Redis saturation appears first, treat the service boundary guidance in Phase 2 as the next action.
+3. If CPU, memory, or Redis saturation appears first, treat the Phase 2 service boundary guidance as the next action.
 4. If the same threshold trips on multiple runs across 7 days, open the Phase 2 or Phase 3 planning path instead of tuning in place.
 
-Decision note:
-
-- Keep the smallest change that restores the target metric and record the reason in the run sheet.
+Decision note: keep the smallest change that restores the target metric and record the reason in the run sheet.
 
 ## 13) Artifact-to-Worksheet Field Mapping
-
-Use this section when filling the WebSocket and Admin API worksheets from helper-generated artifacts.
+Use when filling the WebSocket and Admin API worksheets from helper-generated artifacts.
 
 ### WebSocket worksheet mapping
-
-Artifact sources:
-
-- `artifacts/<timestamp>/ws-chat-report.json`
-- `artifacts/<timestamp>/ws-chat-summary.txt`
+Artifact sources: `artifacts/<timestamp>/ws-chat-report.json`, `artifacts/<timestamp>/ws-chat-summary.txt`.
 
 Field mapping:
+- `Tool used` → `artillery` (or `scripts/loadtest/run-loadtests.sh -s ws`)
+- `Raw results location` → artifact folder path for this run
+- `Concurrency tiers reached` → `ws-chat.yml` phases (`warmup-10`, `sustain-100`, `sustain-500`, `sustain-1000`)
+- `Error rate` → `vusers.failure_rate_percent` from `ws-chat-summary.txt`
+- `Dropped connections` → derive from `vusers.created - vusers.completed` (use `vusers.failed` as failure signal)
+- `p95 message latency` → first latency line under `summary.*` in `ws-chat-summary.txt` containing `p95=...`
+- `p50 message latency` → first latency line under `summary.*` in `ws-chat-summary.txt` containing `median=...`
 
-- Worksheet `Tool used` -> `artillery` (or `run-loadtests.ps1 -Scenario ws`)
-- Worksheet `Raw results location` -> artifact folder path for this run
-- Worksheet `Concurrency tiers reached` -> `ws-chat.yml` phases (`warmup-10`, `sustain-100`, `sustain-500`, `sustain-1000`)
-- Worksheet `Error rate` -> `vusers.failure_rate_percent` from `ws-chat-summary.txt`
-- Worksheet `Dropped connections` -> derive from `vusers.created - vusers.completed` (use `vusers.failed` as failure signal)
-- Worksheet `p95 message latency` -> first available latency line under `summary.*` in `ws-chat-summary.txt` that includes `p95=...`
-- Worksheet `p50 message latency` -> first available latency line under `summary.*` in `ws-chat-summary.txt` that includes `median=...`
-
-Notes:
-
-- The exact `summary.*` metric names in Artillery output can vary by engine/plugins.
-- If no latency summary field is emitted, record this explicitly and use server-side latency metrics as fallback evidence.
+Notes: the exact `summary.*` metric names in Artillery output can vary by engine/plugins; if no latency summary field is emitted, record this explicitly and use server-side latency metrics as fallback evidence.
 
 ### Admin API worksheet mapping
-
-Artifact sources:
-
-- `artifacts/<timestamp>/admin-api-summary.json`
-- `artifacts/<timestamp>/admin-api-output.txt`
-- `artifacts/<timestamp>/admin-api-endpoint-summary.txt`
+Artifact sources: `artifacts/<timestamp>/admin-api-summary.json`, `artifacts/<timestamp>/admin-api-output.txt`, `artifacts/<timestamp>/admin-api-endpoint-summary.txt`.
 
 Field mapping:
+- `Tool used` → `k6` (or `scripts/loadtest/run-loadtests.sh -s admin`)
+- `Raw results location` → artifact folder path for this run
+- `/api/admin/overview|analytics|chat-log|users` p95 → `overview|analytics|chat_log|users` row `P95Ms` in `admin-api-endpoint-summary.txt`
+- `Error rate` → `Global failure rate` in `admin-api-endpoint-summary.txt`
+- `Slowest endpoint` → endpoint row with highest `P95Ms`
 
-- Worksheet `Tool used` -> `k6` (or `run-loadtests.ps1 -Scenario admin`)
-- Worksheet `Raw results location` -> artifact folder path for this run
-- Worksheet `/api/admin/overview` p95 -> `overview` row `P95Ms` in `admin-api-endpoint-summary.txt`
-- Worksheet `/api/admin/analytics` p95 -> `analytics` row `P95Ms` in `admin-api-endpoint-summary.txt`
-- Worksheet `/api/admin/chat-log` p95 -> `chat_log` row `P95Ms` in `admin-api-endpoint-summary.txt`
-- Worksheet `/api/admin/users` p95 -> `users` row `P95Ms` in `admin-api-endpoint-summary.txt`
-- Worksheet `Error rate` -> `Global failure rate` in `admin-api-endpoint-summary.txt`
-- Worksheet `Slowest endpoint` -> endpoint row with highest `P95Ms`
+Notes: keep the raw `admin-api-summary.json` file for audits and re-parsing; if an endpoint row is missing, treat it as a test/data issue and re-run with full endpoint coverage.
 
-Notes:
-
-- Keep the raw `admin-api-summary.json` file for audits and re-parsing.
-- If an endpoint row is missing, treat as test/data issue and re-run with full endpoint coverage.
+Summary helper (bash): run `scripts/loadtest/summarize-artillery.sh <report.json>` after an Artillery WS run, and `scripts/loadtest/summarize-k6.sh <summary.json>` after a k6 admin run, to produce the artifact files used in the mappings above.

frontend/index.html

@@ -44,29 +44,30 @@
           <div class="chat-header-actions">
             <span id="chat-user-label" class="chat-user-label" hidden></span>
             <button id="auth-toggle-btn" type="button" class="btn btn-secondary">Sign in</button>
-            <button id="auth-logout-btn" type="button" class="btn btn-secondary" hidden>Sign out</button> <button id="users-toggle-btn" type="button" class="btn btn-secondary" aria-controls="presence-panel" aria-expanded="true">Users</button> <button type="button" class="btn btn-secondary" onclick="togglePrivateDM()">New DM</button>
+            <button id="auth-logout-btn" type="button" class="btn btn-secondary" hidden>Sign out</button>
+            <button type="button" class="btn btn-secondary" onclick="togglePrivateDM()">New DM</button>
           </div>
         </header>
 
-        <div id="auth-panel" class="auth-panel" hidden>
-          <input id="auth-username" type="text" class="input" placeholder="Username" autocomplete="username" />
-          <input id="auth-password" type="password" class="input" placeholder="Password" autocomplete="current-password" />
-          <div class="button-row">
-            <button id="auth-login-btn" type="button" class="btn btn-primary">Log in</button>
-            <button id="auth-register-btn" type="button" class="btn btn-secondary">Register</button>
+        <div id="auth-panel" class="auth-modal" hidden>
+          <div id="auth-backdrop" class="auth-modal-backdrop"></div>
+          <div class="auth-modal-content">
+            <button id="auth-close-btn" type="button" class="btn auth-close" aria-label="Close sign in">×</button>
+            <h2 class="auth-modal-title">Sign in</h2>
+            <input id="auth-username" type="text" class="input" placeholder="Username" autocomplete="username" />
+            <input id="auth-password" type="password" class="input" placeholder="Password" autocomplete="current-password" />
+            <div class="button-row">
+              <button id="auth-login-btn" type="button" class="btn btn-primary">Log in</button>
+              <button id="auth-register-btn" type="button" class="btn btn-secondary">Register</button>
+            </div>
+            <p id="auth-status" class="status-line" aria-live="polite"></p>
           </div>
-          <p id="auth-status" class="status-line" aria-live="polite"></p>
         </div>
 
         <div id="chat-flash" class="chat-flash" aria-live="polite" hidden></div>
 
         <section id="messages" class="chat-messages" aria-live="polite"></section>
 
-        <div id="dm-search-row" class="dm-search-row" hidden>
-          <input id="dm-search-input" type="text" class="input" placeholder="Search this conversation..." />
-          <button id="dm-search-btn" type="button" class="btn btn-secondary">Search</button>
-        </div>
-
         <section id="presence-panel" class="presence-panel">
           <div id="presence-list" class="presence-list"></div>
         </section>

frontend/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "frontend",
-  "version": "1.0.11",
+  "version": "1.1.0",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "frontend",
-      "version": "1.0.11",
+      "version": "1.1.0",
       "dependencies": {
         "hls.js": "^1.6.16"
       },

frontend/package.json

@@ -1,7 +1,7 @@
 {
   "name": "frontend",
   "private": true,
-  "version": "1.0.13",
+  "version": "1.1.0",
   "type": "module",
   "engines": {
     "node": ">=22.0.0"

frontend/src/admin.ts

@@ -31,6 +31,7 @@ const adminConfig = {
 const MAX_UPLOAD_BYTES = 2 * 1024 * 1024 * 1024;
 const UPLOAD_CHUNK_BYTES = 8 * 1024 * 1024;
 const UPLOAD_RESUME_KEY_PREFIX = "linearvoid-upload-resume:";
+const API_URL = import.meta.env.VITE_PUBLIC_APP_URL?.trim() || window.location.origin;
 
 const state = {
   mediaUploading: false,
@@ -131,7 +132,7 @@ const pageCache: Record<string, string> = {};
 async function loadPageContent(tabName: string): Promise<string> {
   if (pageCache[tabName]) return pageCache[tabName];
 
-  const response = await fetch(`/admin/${tabName}.html`);
+  const response = await fetch(`${API_URL}/admin/${tabName}.html`);
   if (!response.ok) throw new Error(`Failed to load page: ${tabName}`);
   const html = await response.text();
   pageCache[tabName] = html;
@@ -175,7 +176,7 @@ async function switchTab(tabName: string) {
 
 export async function refreshOverview() {
   try {
-    const response = await fetch("/api/admin/overview");
+    const response = await fetch(`${API_URL}/api/admin/overview`);
     if (!response.ok) throw new Error(`HTTP ${response.status}`);
 
     const payload = await response.json();
@@ -202,7 +203,7 @@ export async function refreshOverview() {
 export async function refreshAnalytics() {
   setStatus("analytics-status", "Loading analytics...");
   try {
-    const response = await fetch("/api/admin/analytics");
+    const response = await fetch(`${API_URL}/api/admin/analytics`);
     if (!response.ok) throw new Error(`HTTP ${response.status}`);
 
     const payload = await response.json();
@@ -240,7 +241,7 @@ export async function refreshAnalytics() {
 async function refreshPlaylist() {
   setStatus("playlist-status", "Loading playlist...");
   try {
-    const response = await fetch("/api/admin/playlist");
+    const response = await fetch(`${API_URL}/api/admin/playlist`);
     if (!response.ok) throw new Error(`HTTP ${response.status}`);
 
     const payload = await response.json();
@@ -522,7 +523,7 @@ async function submitOverride(e: Event) {
   const endTime = (document.getElementById("override-end") as HTMLInputElement).value;
   const reason = (document.getElementById("override-reason") as HTMLInputElement).value;
   try {
-    const resp = await fetch("/api/admin/schedule/overrides", {
+    const resp = await fetch(`${API_URL}/api/admin/schedule/overrides`, {
       method: "POST",
       headers: { "Content-Type": "application/json" },
       body: JSON.stringify({
@@ -551,7 +552,7 @@ async function submitOverride(e: Event) {
 
 async function deleteOverride(id: number) {
   try {
-    await fetch(`/api/admin/schedule/overrides?id=${id}`, { method: "DELETE" });
+    await fetch(`${API_URL}/api/admin/schedule/overrides?id=${id}`, { method: "DELETE" });
     refreshOverrides();
   } catch {
     /* ignore */
@@ -561,17 +562,21 @@ async function deleteOverride(id: number) {
 async function renderServiceRow(service: any) {
   const svcDetail = service.detail || "";
 
-  // playout service returns json detail, all others return ISO formatted time string
+  // playout service returns json detail, all others return `local process` or `ping ok`
   if (service.name === "playout" && svcDetail) {
     try {
       // create a list of key-value pairs for each property in the JSON object
       const detailObj = JSON.parse(svcDetail);
-      service.detail = Object.entries(detailObj)
-        .map(([key, value]) => `${key}: ${value}`)
-        .join("\n");
+      // create a table of key-value pairs for each property in the JSON object
+      const detailTable = Object.entries(detailObj)
+        .map(([key, value]) => `<tr><td><strong>${key}</strong></td><td>${value}</td></tr>`)
+        .join("");
+      service.detail = `<p><table class="service-detail-table">${detailTable}</table></p>`;
     } catch {
       // ignore JSON parse errors
     }
+  } else {
+    service.detail = `<p>${service.detail}</p>`;
   }
 
   const row = document.createElement("article");
@@ -583,7 +588,7 @@ async function renderServiceRow(service: any) {
     </div>
     <div class="service-row-right">
       <p class="service-status service-status-${service.status}">${service.status}</p>
-      <p>${service.detail}</p>
+      ${service.detail}
     </div>
   `;
   return row;
@@ -749,9 +754,12 @@ async function uploadVideo(file: File) {
       setStatus("media-status", `Uploading ${file.name} (${chunkIndex + 1}/${effectiveTotalChunks} chunks)...`);
     }
 
-    const completeResponse = await fetch(`/api/admin/media/uploads/${encodeURIComponent(uploadId)}/complete`, {
-      method: "POST",
-    });
+    const completeResponse = await fetch(
+      `${API_URL}/api/admin/media/uploads/${encodeURIComponent(uploadId)}/complete`,
+      {
+        method: "POST",
+      },
+    );
 
     if (!completeResponse.ok) {
       throw new Error(await readErrorMessage(completeResponse));
@@ -791,7 +799,7 @@ async function editMedia(file: MediaEntry) {
 
   setStatus("media-status", `Saving metadata for ${file.name}...`);
   try {
-    const response = await fetch(`/api/admin/media/${file.id}`, {
+    const response = await fetch(`${API_URL}/api/admin/media/${file.id}`, {
       method: "PUT",
       headers: { "Content-Type": "application/json" },
       body: JSON.stringify({ title, status, tags }),
@@ -816,7 +824,7 @@ async function normalizeMedia(file: MediaEntry) {
 
   setStatus("media-status", `Normalizing ${file.name}...`);
   try {
-    const response = await fetch(`/api/admin/media/${file.id}/normalize`, { method: "POST" });
+    const response = await fetch(`${API_URL}/api/admin/media/${file.id}/normalize`, { method: "POST" });
     if (!response.ok) throw new Error(await readErrorMessage(response));
 
     setStatus("media-status", `Normalization started for ${file.name}.`);
@@ -836,7 +844,7 @@ function pollNormalizationStatus(id: number, attempt: number) {
   }
   window.setTimeout(async () => {
     try {
-      const response = await fetch(`/api/admin/media/${id}`);
+      const response = await fetch(`${API_URL}/api/admin/media/${id}`);
       if (!response.ok) return;
       const file = (await response.json()) as MediaEntry;
       if (file.status === "normalizing" || file.status === "needs_normalization") {
@@ -869,7 +877,7 @@ async function deleteMedia(file: MediaEntry) {
 
   setStatus("media-status", `Deleting ${file.name}...`);
   try {
-    const response = await fetch(`/api/admin/media/${file.id}`, {
+    const response = await fetch(`${API_URL}/api/admin/media/${file.id}`, {
       method: "DELETE",
     });
     if (!response.ok) throw new Error(`HTTP ${response.status}`);
@@ -887,7 +895,7 @@ async function deleteMedia(file: MediaEntry) {
 async function refreshModerationLogs() {
   setStatus("moderation-status", "Loading moderation logs...");
   try {
-    const response = await fetch(`/api/admin/moderation/logs?limit=${adminConfig.moderationLogLimit}`);
+    const response = await fetch(`${API_URL}/api/admin/moderation/logs?limit=${adminConfig.moderationLogLimit}`);
     if (!response.ok) throw new Error(`HTTP ${response.status}`);
 
     const payload = await response.json();
@@ -988,7 +996,7 @@ async function refreshChatLog() {
     if (chatLogState.username) params.set("username", chatLogState.username);
     if (chatLogState.text) params.set("text", chatLogState.text);
 
-    const response = await fetch(`/api/admin/chat-log?${params}`);
+    const response = await fetch(`${API_URL}/api/admin/chat-log?${params}`);
     if (!response.ok) throw new Error(`HTTP ${response.status}`);
 
     const payload = await response.json();
@@ -1118,7 +1126,7 @@ async function deleteChatEntry(id: number) {
 
   setStatus("chat-log-status", `Deleting message #${id}...`);
   try {
-    const response = await fetch(`/api/admin/chat-log/${id}`, {
+    const response = await fetch(`${API_URL}/api/admin/chat-log/${id}`, {
       method: "DELETE",
     });
     if (!response.ok) throw new Error(`HTTP ${response.status}`);
@@ -1268,7 +1276,7 @@ async function restartService(service: string) {
   const status = document.getElementById("restart-status");
   if (status) status.textContent = `Requesting restart of ${service}...`;
   try {
-    const resp = await fetch(`/api/admin/restart?service=${service}`, { method: "POST" });
+    const resp = await fetch(`${API_URL}/api/admin/restart?service=${service}`, { method: "POST" });
     const data = await resp.json();
     if (status) status.textContent = data.message;
   } catch (err) {

frontend/src/main.test.ts

@@ -124,19 +124,18 @@ beforeEach(() => {
     <div id="chat-user-label" hidden></div>
     <button id="auth-toggle-btn" type="button">Sign in</button>
     <button id="auth-logout-btn" type="button" hidden>Logout</button>
-    <div id="auth-panel" hidden>
-      <input id="auth-username" type="text" />
-      <input id="auth-password" type="password" />
-      <span id="auth-status"></span>
-      <button id="auth-login-btn" type="button">Login</button>
-      <button id="auth-register-btn" type="button">Register</button>
+    <div id="auth-panel" class="auth-modal" hidden>
+      <div id="auth-backdrop" class="auth-modal-backdrop"></div>
+      <div class="auth-modal-content">
+        <button id="auth-close-btn" type="button">×</button>
+        <input id="auth-username" type="text" />
+        <input id="auth-password" type="password" />
+        <span id="auth-status"></span>
+        <button id="auth-login-btn" type="button">Login</button>
+        <button id="auth-register-btn" type="button">Register</button>
+      </div>
     </div>
-    <button id="users-toggle-btn" type="button" aria-expanded="true">Users</button>
     <div id="typing-indicator" hidden></div>
-    <div id="dm-search-row" hidden>
-      <input id="dm-search-input" type="text" />
-      <button id="dm-search-btn" type="button">Search</button>
-    </div>
   `;
 
   // Stub prompt for DM
@@ -253,6 +252,25 @@ describe("WebSocket reconnect", () => {
     expect(websocketInstances).toHaveLength(2);
     expect(websocketInstances[1].url).toContain("room=lobby");
   });
+
+  it("does not reconnect from a socket that was intentionally replaced", async () => {
+    vi.useFakeTimers();
+
+    window.switchRoom("lobby");
+    expect(websocketInstances).toHaveLength(1);
+    const replaced = websocketInstances[0];
+
+    // A room switch intentionally replaces the socket; the old socket must not
+    // schedule its own reconnect, otherwise history is re-pulled on a loop.
+    window.switchRoom("main");
+    expect(websocketInstances).toHaveLength(2);
+    expect(replaced.onclose).toBeNull();
+
+    replaced.onclose?.();
+    await vi.advanceTimersByTimeAsync(3000);
+
+    expect(websocketInstances).toHaveLength(2);
+  });
 });
 
 describe("URL resolution", () => {
@@ -437,7 +455,7 @@ describe("sendMessage over websocket", () => {
   });
 });
 
-describe("auth and user list wiring", () => {
+describe("auth wiring", () => {
   it("logs in and updates the auth UI", async () => {
     await import("../src/main");
 
@@ -467,30 +485,6 @@ describe("auth and user list wiring", () => {
     expect(document.getElementById("auth-status")!.textContent).toContain("Signed in as alice");
     expect(document.getElementById("chat-user-label")!.textContent).toBe("alice");
   });
-
-  it("toggles the user list visibility", async () => {
-    await import("../src/main");
-
-    globalThis.fetch = vi.fn().mockResolvedValue({
-      ok: true,
-      json: async () => ({ authenticated: false, entries: [] }),
-      text: async () => "",
-    }) as any;
-
-    document.dispatchEvent(new Event("DOMContentLoaded"));
-
-    const panel = document.getElementById("presence-panel")!;
-    const btn = document.getElementById("users-toggle-btn")!;
-    expect(panel.hidden).toBe(false);
-
-    btn.click();
-    expect(panel.hidden).toBe(true);
-    expect(btn.getAttribute("aria-expanded")).toBe("false");
-
-    btn.click();
-    expect(panel.hidden).toBe(false);
-    expect(btn.getAttribute("aria-expanded")).toBe("true");
-  });
 });
 
 describe("player error recovery", () => {

frontend/src/main.ts

@@ -16,9 +16,7 @@ let ws: WebSocket | null = null;
 let currentRoom = "main";
 let username = "viewer" + Math.floor(Math.random() * 1000);
 let authenticated = false;
-let currentDMTarget = "";
 let chatVisible = true;
-let usersVisible = true;
 let typingUser = "";
 let lastTypingSent = 0;
 let typingTimeout: number | undefined;
@@ -38,7 +36,6 @@ export function resetViewerStateForTests() {
   username = "viewer";
   authenticated = false;
   chatVisible = true;
-  usersVisible = true;
   typingUser = "";
   lastTypingSent = 0;
   onlineUsers.clear();
@@ -79,19 +76,32 @@ export function resolveManifestUrl() {
 }
 
 function connectWebSocket(room: string) {
-  if (ws) ws.close();
+  // Clear previous messages so reconnects render a fresh history instead of
+  // appending duplicates on top of the old log.
+  clearMessages();
+
+  // Detach the old socket's close handler before closing it. Otherwise an
+  // intentional reconnect (login/logout/room switch/session refresh) would
+  // have the replaced socket schedule yet another reconnect, and each new
+  // connection re-pulls the room's history — filling the panel with old
+  // messages on a loop.
+  if (ws) {
+    ws.onclose = null;
+    ws.close();
+  }
 
   const socketUrl = new URL(resolveWebSocketBaseUrl());
   socketUrl.searchParams.set("username", username);
   socketUrl.searchParams.set("room", room);
-  ws = new WebSocket(socketUrl);
+  const socket = new WebSocket(socketUrl);
+  ws = socket;
 
-  ws.onopen = () => {
+  socket.onopen = () => {
     console.log(`Connected to chat room: ${room}`);
     addMessage("system", "Connected to " + room);
   };
 
-  ws.onmessage = (event) => {
+  socket.onmessage = (event) => {
     const msg = JSON.parse(event.data);
     if (msg.type === "typing") {
       showTyping(msg.username);
@@ -119,12 +129,20 @@ function connectWebSocket(room: string) {
     addMessage(msg.type || "public", msg.text, msg.username, msg.timestamp);
   };
 
-  ws.onclose = () => {
+  socket.onclose = () => {
+    // Ignore close events from sockets that were intentionally replaced.
+    if (ws !== socket) return;
+    ws = null;
     console.log("Chat disconnected. Reconnecting in 3s...");
     setTimeout(() => connectWebSocket(room), 3000);
   };
 }
 
+function clearMessages() {
+  const messages = document.getElementById("messages");
+  if (messages) messages.innerHTML = "";
+}
+
 function addMessage(type: string, text: string, sender = "", timestamp = new Date()) {
   const messages = document.getElementById("messages");
   if (!messages) return;
@@ -177,7 +195,8 @@ function renderPresence() {
 
 async function refreshSchedule() {
   try {
-    const res = await fetch("/api/schedule");
+    const apiUrl = import.meta.env.VITE_PUBLIC_APP_URL?.trim() || window.location.origin;
+    const res = await fetch(`${apiUrl}/api/schedule`);
     if (!res.ok) return;
     const data = await res.json();
     const entries = data.entries as Array<{ title: string; status: string }>;
@@ -268,6 +287,9 @@ export async function initPlayerAndChat(loadHlsModule: () => Promise<{ default:
     }
   }
 
+  // Resolve the user session first so we connect once with the correct username
+  // (avoids pushing chat history twice on initial load).
+  await sessionReady;
   connectWebSocket("main");
 
   // Start schedule polling
@@ -324,46 +346,34 @@ function sendMessage() {
 
 function switchRoom(room: string) {
   currentRoom = room;
-  const isDM = room.startsWith("dm:");
   document.getElementById("room-name")!.textContent =
     room === "main" ? "Main Channel" : room.charAt(0).toUpperCase() + room.slice(1) + " Room";
-  addMessage("system", `Switched to ${room} room`);
-  const searchRow = document.getElementById("dm-search-row");
-  if (searchRow) searchRow.hidden = !isDM;
   connectWebSocket(room);
 }
 
 function togglePrivateDM() {
   const target = prompt("Enter username to start private chat with:");
   if (target && target !== username) {
-    currentDMTarget = target;
     const dmRoom = `dm:${[username, target].sort().join("-")}`;
     switchRoom(dmRoom);
   }
 }
 
-async function searchDM() {
-  const input = document.getElementById("dm-search-input") as HTMLInputElement;
-  const q = input.value.trim();
-  if (!q || !currentDMTarget) return;
-  try {
-    const res = await fetch(`/api/dm/search?with=${encodeURIComponent(currentDMTarget)}&q=${encodeURIComponent(q)}`);
-    if (!res.ok) return;
-    const results = await res.json();
-    addMessage("system", `Search results (${results.length}) for "${q}":`);
-    for (const m of results) {
-      addMessage(m.type || "private", m.text, m.username, m.timestamp);
-    }
-  } catch {
-    /* ignore */
-  }
-}
-
 window.switchRoom = switchRoom;
 window.sendMessage = sendMessage;
 window.togglePrivateDM = togglePrivateDM;
 window.toggleChatPanel = toggleChatPanel;
 
+function openAuthModal() {
+  const panel = document.getElementById("auth-panel");
+  if (panel) panel.hidden = false;
+}
+
+function closeAuthModal() {
+  const panel = document.getElementById("auth-panel");
+  if (panel) panel.hidden = true;
+}
+
 function updateAuthUI() {
   const label = document.getElementById("chat-user-label");
   const toggle = document.getElementById("auth-toggle-btn");
@@ -378,24 +388,26 @@ function updateAuthUI() {
   if (panel) panel.hidden = true;
 }
 
-async function refreshUserSession() {
-  try {
-    const res = await fetch("/api/auth/session");
-    if (!res.ok) return;
-    const data = await res.json();
-    const wasAnonymous = username.startsWith("viewer");
-    if (data.authenticated && data.username) {
-      authenticated = true;
-      username = data.username;
-      updateAuthUI();
-      if (wasAnonymous) connectWebSocket(currentRoom);
-    } else {
-      authenticated = false;
+let sessionReady: Promise<void> = Promise.resolve();
+
+function refreshUserSession(): void {
+  sessionReady = (async () => {
+    try {
+      const apiUrl = import.meta.env.VITE_PUBLIC_APP_URL?.trim() || window.location.origin;
+      const res = await fetch(`${apiUrl}/api/auth/session`, { credentials: "include" });
+      if (!res.ok) return;
+      const data = await res.json();
+      if (data.authenticated && data.username) {
+        authenticated = true;
+        username = data.username;
+      } else {
+        authenticated = false;
+      }
       updateAuthUI();
+    } catch {
+      /* non-critical */
     }
-  } catch {
-    /* non-critical */
-  }
+  })();
 }
 
 async function submitAuth(path: string) {
@@ -410,7 +422,8 @@ async function submitAuth(path: string) {
     return;
   }
   try {
-    const res = await fetch(path, {
+    const apiUrl = import.meta.env.VITE_PUBLIC_APP_URL?.trim() || window.location.origin;
+    const res = await fetch(`${apiUrl}${path}`, {
       method: "POST",
       headers: { "Content-Type": "application/json" },
       body: JSON.stringify({ username: u, password: p }),
@@ -446,7 +459,8 @@ function register() {
 
 async function logout() {
   try {
-    await fetch("/api/auth/logout", { method: "POST" });
+    const apiUrl = import.meta.env.VITE_PUBLIC_APP_URL?.trim() || window.location.origin;
+    await fetch(`${apiUrl}/api/auth/logout`, { method: "POST", credentials: "include" });
   } catch {
     /* ignore */
   }
@@ -456,15 +470,6 @@ async function logout() {
   connectWebSocket(currentRoom);
 }
 
-function toggleUserList() {
-  const panel = document.getElementById("presence-panel");
-  const btn = document.getElementById("users-toggle-btn");
-  if (!panel) return;
-  usersVisible = !usersVisible;
-  panel.hidden = !usersVisible;
-  if (btn) btn.setAttribute("aria-expanded", String(usersVisible));
-}
-
 function showTyping(name: string) {
   if (name === username) return;
   const el = document.getElementById("typing-indicator");
@@ -501,15 +506,14 @@ function sendTyping() {
 
 document.addEventListener("DOMContentLoaded", () => {
   void initPlayerAndChat();
-  void refreshUserSession();
+  refreshUserSession();
   document.getElementById("auth-toggle-btn")?.addEventListener("click", () => {
-    const panel = document.getElementById("auth-panel");
-    if (panel) panel.hidden = !panel.hidden;
+    if (!authenticated) openAuthModal();
   });
+  document.getElementById("auth-close-btn")?.addEventListener("click", closeAuthModal);
+  document.getElementById("auth-backdrop")?.addEventListener("click", closeAuthModal);
   document.getElementById("auth-login-btn")?.addEventListener("click", login);
   document.getElementById("auth-register-btn")?.addEventListener("click", register);
   document.getElementById("auth-logout-btn")?.addEventListener("click", logout);
-  document.getElementById("dm-search-btn")?.addEventListener("click", searchDM);
-  document.getElementById("users-toggle-btn")?.addEventListener("click", toggleUserList);
   document.getElementById("message-input")?.addEventListener("input", sendTyping);
 });

frontend/src/style.css

@@ -305,19 +305,53 @@ select {
   white-space: nowrap;
 }
 
-.auth-panel {
+.auth-modal {
+  position: fixed;
+  inset: 0;
+  z-index: 100;
+  display: flex;
+  align-items: center;
+  justify-content: center;
+}
+
+.auth-modal[hidden] {
+  display: none;
+}
+
+.auth-modal-backdrop {
+  position: absolute;
+  inset: 0;
+  background: rgba(0, 0, 0, 0.6);
+}
+
+.auth-modal-content {
+  position: relative;
+  z-index: 1;
   display: flex;
   flex-direction: column;
   gap: 0.5rem;
-  padding: 0.75rem;
-  border-top: 1px solid var(--bg-2);
+  width: min(20rem, calc(100vw - 2rem));
+  padding: 1.25rem;
+  border: 1px solid var(--bg-2);
+  border-radius: 0.5rem;
+  background: var(--panel-2);
+  box-shadow: 0 8px 30px rgba(0, 0, 0, 0.4);
 }
 
-.dm-search-row {
-  display: flex;
-  gap: 0.5rem;
-  padding: 0.5rem 0.75rem;
-  border-top: 1px solid var(--bg-2);
+.auth-close {
+  position: absolute;
+  top: 0.5rem;
+  right: 0.5rem;
+  padding: 0.25rem 0.5rem;
+  font-size: 1.2rem;
+  line-height: 1;
+}
+
+.auth-modal-title {
+  margin: 0;
+  font-size: 1rem;
+  font-weight: 600;
+  color: var(--ink-0);
 }
 
 .typing-indicator {
@@ -1010,6 +1044,12 @@ select {
   text-align: right;
 }
 
+.service-detail-table {
+  width: 100%;
+  border-collapse: collapse;
+  text-align: left;
+}
+
 .service-status {
   margin: 0;
   font-weight: 700;

frontend/wrangler.toml

@@ -0,0 +1,6 @@
+name = "linearvoid-frontend"
+compatibility_date = "2025-08-01"
+
+[assets]
+directory = "./dist"
+not_found_handling = "single-page-application"

infrastructure/cloudflare/.gitignore

@@ -0,0 +1,5 @@
+node_modules/
+.wrangler/
+dist/
+coverage/
+*.log

infrastructure/cloudflare/README.md

@@ -0,0 +1,65 @@
+# linearvoid-media Worker
+
+Cloudflare Worker that owns **`media.linearvoid.tv`** and serves the HLS objects
+stored in the `linearvoid-media` R2 bucket, applying the edge cache policy in code:
+
+| Object                                      | Cache-Control                                    |
+| ------------------------------------------- | ------------------------------------------------ |
+| `.m4s` / `.ts` / `.mp4` / `.m4a` (segments) | `public, max-age=31536000, immutable`            |
+| `.m3u8` (live manifest)                     | `public, max-age=1, s-maxage=2, must-revalidate` |
+
+It also adds HLS CORS (`Access-Control-Allow-Origin: *`), OPTIONS preflight, a
+`/health` probe, and Range-request support. This worker replaces direct R2
+custom-domain serving on the media host so cache headers/CORS are controlled in
+code (cache rules on the same host remain the origin baseline).
+
+## Layout
+
+```text
+infrastructure/cloudflare/
+├── wrangler.toml        # name/binding + production custom-domain route
+├── package.json
+├── src/worker.js        # worker module (exports default.fetch + pure helpers)
+└── test/                # Vitest unit tests + Miniflare (workerd) R2 tests
+```
+
+## Local dev / test
+
+```bash
+cd infrastructure/cloudflare
+npm install
+npm test                 # unit + Miniflare R2 integration tests
+npx wrangler dev         # run locally against your account's R2 bucket
+```
+
+## Deploy
+
+**Preview / test target** (default environment → `*.workers.dev`):
+
+```bash
+cd infrastructure/cloudflare
+npx wrangler deploy
+```
+
+**Production** (attaches `media.linearvoid.tv` as a custom domain on this worker):
+
+```bash
+cd infrastructure/cloudflare
+npx wrangler deploy --env production
+```
+
+Both need `CLOUDFLARE_API_TOKEN` (Workers Scripts/Routes edit) and
+`CLOUDFLARE_ACCOUNT_ID` in the environment. The media host must first be removed
+from the R2 bucket's direct custom-domain binding, otherwise the route collides.
+
+Automated production deploys run via
+[`.gitea/workflows/workers-media-deploy.yml`](../../.gitea/workflows/workers-media-deploy.yml)
+after CI succeeds on `main`.
+
+## Verify
+
+```bash
+curl -I https://media.linearvoid.tv/hls/playlist.m3u8   # short TTL
+curl -I https://media.linearvoid.tv/hls/<segment>.m4s   # immutable + CORS
+curl https://media.linearvoid.tv/health                 # {"status":"ok",...}
+```

infrastructure/cloudflare/package.json

@@ -0,0 +1,20 @@
+{
+  "name": "linearvoid-media-worker",
+  "version": "1.0.0",
+  "private": true,
+  "description": "Linear Void media edge worker: serves HLS from R2 on media.linearvoid.tv with immutable segment caching.",
+  "type": "module",
+  "scripts": {
+    "dev": "wrangler dev",
+    "start": "wrangler dev",
+    "deploy": "wrangler deploy",
+    "deploy:production": "wrangler deploy --env production",
+    "test": "vitest run",
+    "test:watch": "vitest"
+  },
+  "devDependencies": {
+    "miniflare": "^4.0.0",
+    "vitest": "^4.1.10",
+    "wrangler": "^4.0.0"
+  }
+}

infrastructure/cloudflare/src/worker.js

@@ -0,0 +1,144 @@
+// Linear Void media edge worker (linearvoid-media).
+//
+// Owns the media.linearvoid.tv host and serves HLS objects from the R2 bucket
+// bound as MEDIA_BUCKET. Its purpose (arc42 quality goal #1, cost efficiency) is
+// to guarantee immutable, long-lived caching of media segments at the edge while
+// keeping live manifests fresh — the caching baseline is also configured as
+// Cloudflare cache rules on the same host, but this worker enforces it in code.
+//
+// Cache policy (matches docs/infrastructure/cloudflare/cache-rules.md + STREAMING.md):
+//   .m4s / .ts / .mp4 / .m4a  -> public, max-age=31536000, immutable
+//   .m3u8 (live manifest)      -> public, max-age=1, s-maxage=2, must-revalidate
+
+const IMMUTABLE_CACHE = "public, max-age=31536000, immutable";
+const MANIFEST_CACHE = "public, max-age=1, s-maxage=2, must-revalidate";
+const DEFAULT_CACHE = "public, max-age=60";
+
+const IMMUTABLE_SUFFIXES = [".m4s", ".ts", ".mp4", ".m4a"];
+const MANIFEST_SUFFIXES = [".m3u8"];
+
+const CONTENT_TYPES = {
+  ".m3u8": "application/vnd.apple.mpegurl",
+  ".m4s": "video/mp4",
+  ".mp4": "video/mp4",
+  ".ts": "video/MP2T",
+  ".m4a": "audio/mp4",
+};
+
+/** Map a request path to its edge Cache-Control value. Pure + testable. */
+export function cacheControlForPath(pathname) {
+  const lower = pathname.toLowerCase();
+  for (const suffix of MANIFEST_SUFFIXES) {
+    if (lower.endsWith(suffix)) return MANIFEST_CACHE;
+  }
+  for (const suffix of IMMUTABLE_SUFFIXES) {
+    if (lower.endsWith(suffix)) return IMMUTABLE_CACHE;
+  }
+  return DEFAULT_CACHE;
+}
+
+/** Map a request path to a browser-friendly Content-Type. Pure + testable. */
+export function contentTypeForPath(pathname) {
+  const lower = pathname.toLowerCase();
+  for (const [suffix, type] of Object.entries(CONTENT_TYPES)) {
+    if (lower.endsWith(suffix)) return type;
+  }
+  return "application/octet-stream";
+}
+
+const CORS_HEADERS = {
+  "Access-Control-Allow-Origin": "*",
+  "Access-Control-Allow-Methods": "GET, HEAD, OPTIONS",
+  "Access-Control-Allow-Headers": "Range, If-None-Match, If-Modified-Since",
+  "Access-Control-Max-Age": "86400",
+};
+
+function json(body, status = 200) {
+  return new Response(JSON.stringify(body), {
+    status,
+    headers: { "Content-Type": "application/json", ...CORS_HEADERS },
+  });
+}
+
+// Parse a single, closed byte range ("bytes=start-end") into an R2 range object.
+// Returns null for open/suffix/multi/invalid ranges — the caller then serves the
+// full object (HTTP 200), which is valid per RFC 7233 (a server may ignore Range).
+export function parseByteRange(header) {
+  if (!header) return null;
+  const m = /^bytes=(\d*)-(\d*)$/.exec(header.trim());
+  if (!m) return null;
+  const start = m[1];
+  const end = m[2];
+  if (start === "" || end === "") return null; // open or suffix range
+  const offset = Number(start);
+  const length = Number(end) - offset + 1;
+  if (!Number.isInteger(offset) || !Number.isInteger(length) || offset < 0 || length < 1) {
+    return null;
+  }
+  return { offset, length };
+}
+
+export default {
+  async fetch(request, env) {
+    const url = new URL(request.url);
+
+    // CORS preflight for HLS media.
+    if (request.method === "OPTIONS") {
+      return new Response(null, { status: 204, headers: CORS_HEADERS });
+    }
+
+    // Liveness probe used by uptime/monitoring (see cloudflare-uptimerobot-checklist).
+    if (url.pathname === "/health") {
+      return json({ status: "ok", service: "linearvoid-media" });
+    }
+
+    if (request.method !== "GET" && request.method !== "HEAD") {
+      return json({ error: "method not allowed" }, 405);
+    }
+
+    const key = decodeURIComponent(url.pathname).replace(/^\/+/, "");
+    if (!key) {
+      return json({ error: "not found" }, 404);
+    }
+
+    if (!env.MEDIA_BUCKET) {
+      return json({ error: "media bucket not configured" }, 500);
+    }
+
+    // Serve from R2, honouring single closed Range requests (used by some hls.js
+    // / player paths for byte-range fetches). Unsupported ranges fall back to a
+    // full-body 200, which clients accept.
+    const rangeHeader = request.headers.get("range");
+    const parsedRange = parseByteRange(rangeHeader);
+    const object = parsedRange
+      ? await env.MEDIA_BUCKET.get(key, { range: parsedRange })
+      : await env.MEDIA_BUCKET.get(key);
+
+    if (object === null) {
+      return json({ error: "not found" }, 404);
+    }
+
+    const headers = new Headers();
+    object.writeHttpMetadata(headers);
+    headers.set("Cache-Control", cacheControlForPath(url.pathname));
+    headers.set("Content-Type", contentTypeForPath(url.pathname));
+    headers.set("ETag", object.httpEtag);
+    headers.set("Accept-Ranges", "bytes");
+    headers.set("Access-Control-Allow-Origin", "*");
+
+    let status = 200;
+    if (parsedRange && object.range) {
+      status = 206;
+      const { offset, length } = object.range;
+      headers.set(
+        "Content-Range",
+        `bytes ${offset}-${offset + length - 1}/${object.size}`
+      );
+    }
+
+    return new Response(request.method === "HEAD" ? null : object.body, {
+      status,
+      headers,
+    });
+  },
+};

infrastructure/cloudflare/test/worker.r2.test.js

@@ -0,0 +1,77 @@
+import { Miniflare } from "miniflare";
+import { beforeAll, afterAll, describe, it, expect } from "vitest";
+
+const IMMUTABLE = "public, max-age=31536000, immutable";
+const MANIFEST = "public, max-age=1, s-maxage=2, must-revalidate";
+
+// Real-workerd integration test: boots the worker with an in-memory R2 bucket and
+// dispatches requests through the fetch handler.
+let mf;
+
+beforeAll(async () => {
+  mf = new Miniflare({
+    modules: true,
+    scriptPath: "src/worker.js",
+    r2Buckets: { MEDIA_BUCKET: "linearvoid-media" },
+  });
+  const bucket = await mf.getR2Bucket("MEDIA_BUCKET");
+  await bucket.put("hls/playlist.m3u8", "#EXTM3U\n#EXT-X-ENDLIST\n", {
+    httpMetadata: { contentType: "application/vnd.apple.mpegurl" },
+  });
+  await bucket.put("hls/seg-000001.m4s", new Uint8Array([0, 1, 2, 3, 4]));
+  await bucket.put("hls/init.mp4", new Uint8Array([9, 9, 9]));
+});
+
+afterAll(async () => {
+  await mf?.dispose();
+});
+
+const base = "http://media.linearvoid.tv";
+
+describe("linearvoid-media worker (R2)", () => {
+  it("serves a media segment with immutable cache headers", async () => {
+    const res = await mf.dispatchFetch(`${base}/hls/seg-000001.m4s`);
+    expect(res.status).toBe(200);
+    expect(res.headers.get("cache-control")).toBe(IMMUTABLE);
+    expect(res.headers.get("content-type")).toBe("video/mp4");
+    expect(res.headers.get("access-control-allow-origin")).toBe("*");
+    expect(await res.arrayBuffer()).toEqual(new Uint8Array([0, 1, 2, 3, 4]).buffer);
+  });
+
+  it("serves a live manifest with a short TTL", async () => {
+    const res = await mf.dispatchFetch(`${base}/hls/playlist.m3u8`);
+    expect(res.status).toBe(200);
+    expect(res.headers.get("cache-control")).toBe(MANIFEST);
+    expect(res.headers.get("content-type")).toBe("application/vnd.apple.mpegurl");
+  });
+
+  it("honours Range requests with 206 + Content-Range", async () => {
+    const res = await mf.dispatchFetch(`${base}/hls/seg-000001.m4s`, {
+      headers: { Range: "bytes=0-2" },
+    });
+    expect(res.status).toBe(206);
+    expect(res.headers.get("content-range")).toBe("bytes 0-2/5");
+    const bytes = new Uint8Array(await res.arrayBuffer());
+    expect([...bytes]).toEqual([0, 1, 2]);
+  });
+
+  it("returns 404 for a missing object", async () => {
+    const res = await mf.dispatchFetch(`${base}/hls/does-not-exist.m4s`);
+    expect(res.status).toBe(404);
+  });
+
+  it("answers OPTIONS preflight with CORS headers", async () => {
+    const res = await mf.dispatchFetch(`${base}/hls/seg-000001.m4s`, {
+      method: "OPTIONS",
+    });
+    expect(res.status).toBe(204);
+    expect(res.headers.get("access-control-allow-origin")).toBe("*");
+  });
+
+  it("rejects non-GET/HEAD methods", async () => {
+    const res = await mf.dispatchFetch(`${base}/hls/seg-000001.m4s`, {
+      method: "POST",
+    });
+    expect(res.status).toBe(405);
+  });
+});

infrastructure/cloudflare/test/worker.unit.test.js

@@ -0,0 +1,54 @@
+import { describe, it, expect } from "vitest";
+import { cacheControlForPath, contentTypeForPath, parseByteRange } from "../src/worker.js";
+
+const IMMUTABLE = "public, max-age=31536000, immutable";
+const MANIFEST = "public, max-age=1, s-maxage=2, must-revalidate";
+const DEFAULT = "public, max-age=60";
+
+describe("cacheControlForPath", () => {
+  it("marks media segments as immutable", () => {
+    expect(cacheControlForPath("/hls/seg-000001.m4s")).toBe(IMMUTABLE);
+    expect(cacheControlForPath("/hls/init.mp4")).toBe(IMMUTABLE);
+    expect(cacheControlForPath("/hls/seg-000001.ts")).toBe(IMMUTABLE);
+    expect(cacheControlForPath("/media/audio.m4a")).toBe(IMMUTABLE);
+  });
+
+  it("keeps live manifests short-lived", () => {
+    expect(cacheControlForPath("/hls/playlist.m3u8")).toBe(MANIFEST);
+    expect(cacheControlForPath("/hls/master.m3u8")).toBe(MANIFEST);
+  });
+
+  it("falls back to a sane default for other paths", () => {
+    expect(cacheControlForPath("/hls/unknown.txt")).toBe(DEFAULT);
+  });
+});
+
+describe("contentTypeForPath", () => {
+  it("maps known media types", () => {
+    expect(contentTypeForPath("/hls/playlist.m3u8")).toBe("application/vnd.apple.mpegurl");
+    expect(contentTypeForPath("/hls/seg-000001.m4s")).toBe("video/mp4");
+    expect(contentTypeForPath("/hls/init.mp4")).toBe("video/mp4");
+    expect(contentTypeForPath("/hls/seg-000001.ts")).toBe("video/MP2T");
+    expect(contentTypeForPath("/media/audio.m4a")).toBe("audio/mp4");
+  });
+
+  it("falls back to octet-stream", () => {
+    expect(contentTypeForPath("/hls/unknown.bin")).toBe("application/octet-stream");
+  });
+});
+
+describe("parseByteRange", () => {
+  it("parses a closed byte range", () => {
+    expect(parseByteRange("bytes=0-2")).toEqual({ offset: 0, length: 3 });
+    expect(parseByteRange("bytes=10-19")).toEqual({ offset: 10, length: 10 });
+  });
+
+  it("rejects open, suffix, multi, or malformed ranges", () => {
+    expect(parseByteRange("bytes=0-")).toBeNull();
+    expect(parseByteRange("bytes=-500")).toBeNull();
+    expect(parseByteRange("bytes=0-1,3-4")).toBeNull();
+    expect(parseByteRange("items=0-2")).toBeNull();
+    expect(parseByteRange("")).toBeNull();
+    expect(parseByteRange(null)).toBeNull();
+  });
+});

infrastructure/cloudflare/worker.js

@@ -1,21 +0,0 @@
-// Cloudflare Worker for advanced cache rules and WAF (ADR-001)
-addEventListener("fetch", (event) => {
-  event.respondWith(handleRequest(event.request));
-});
-
-async function handleRequest(request) {
-  const url = new URL(request.url);
-
-  // Add immutable cache headers for .m4s segments
-  if (url.pathname.endsWith(".m4s")) {
-    const response = await fetch(request);
-    const headers = new Headers(response.headers);
-    headers.set("Cache-Control", "public, max-age=31536000, immutable");
-    return new Response(response.body, {
-      status: response.status,
-      headers,
-    });
-  }
-
-  return fetch(request);
-}

infrastructure/cloudflare/wrangler.toml

@@ -0,0 +1,31 @@
+# Cloudflare Worker: linearvoid-media (edge HLS delivery)
+# Serves HLS objects from the `linearvoid-media` R2 bucket on media.linearvoid.tv.
+# Applies immutable caching to media segments and a short TTL to live manifests.
+# This worker replaces direct R2 custom-domain serving on the media host so cache
+# headers/CORS are controlled in code (cache rules remain as the origin baseline).
+
+name = "linearvoid-media"
+main = "src/worker.js"
+compatibility_date = "2025-08-01"
+
+# R2 bucket holding the HLS objects. The request path is mirrored onto the bucket,
+# e.g. GET /hls/playlist.m3u8 -> object key `hls/playlist.m3u8`.
+[[r2_buckets]]
+binding = "MEDIA_BUCKET"
+bucket_name = "linearvoid-media"
+
+# Default environment: deploy to the preview *.workers.dev URL (test target).
+#   npx wrangler deploy
+#
+# Production environment: attach the media host as a custom domain on this worker.
+# Note: bindings are NOT inherited by environments, so r2_buckets is redeclared here.
+#   npx wrangler deploy --env production
+[env.production]
+routes = [
+  { pattern = "media.linearvoid.tv", custom_domain = true }
+]
+
+[[env.production.r2_buckets]]
+binding = "MEDIA_BUCKET"
+bucket_name = "linearvoid-media"
+

infrastructure/nginx/nginx.conf

@@ -8,11 +8,21 @@ http {
 
     resolver 127.0.0.11 ipv6=off valid=30s;
 
+    map_hash_bucket_size 128;
+
     map $http_upgrade $connection_upgrade {
         default upgrade;
         '' close;
     }
 
+    map $http_origin $cors_origin {
+        default "";
+        "https://linearvoid.tv"                              $http_origin;
+        "https://www.linearvoid.tv"                          $http_origin;
+        "https://linearvoid-frontend.zwitschi82.workers.dev" $http_origin;
+        "~^https://[a-zA-Z0-9.-]+\.linearvoid\.tv$"          $http_origin;
+    }
+
     server {
         listen 80;
         server_name www.linearvoid.tv;
@@ -85,6 +95,14 @@ http {
         }
 
         location = /api/admin/login {
+            if ($request_method = OPTIONS) {
+                add_header Access-Control-Allow-Origin $cors_origin;
+                add_header Access-Control-Allow-Methods "GET, POST, PUT, DELETE, OPTIONS";
+                add_header Access-Control-Allow-Headers "Content-Type, Authorization";
+                add_header Access-Control-Allow-Credentials true;
+                add_header Access-Control-Max-Age 86400;
+                return 204;
+            }
             set $chat_upstream http://chat:3000;
             proxy_pass $chat_upstream;
             proxy_http_version 1.1;
@@ -93,9 +111,19 @@ http {
             proxy_set_header X-Forwarded-Proto $scheme;
             proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
             proxy_set_header Cookie $http_cookie;
+            add_header Access-Control-Allow-Origin $cors_origin always;
+            add_header Access-Control-Allow-Credentials true always;
         }
 
         location = /api/admin/logout {
+            if ($request_method = OPTIONS) {
+                add_header Access-Control-Allow-Origin $cors_origin;
+                add_header Access-Control-Allow-Methods "GET, POST, PUT, DELETE, OPTIONS";
+                add_header Access-Control-Allow-Headers "Content-Type, Authorization";
+                add_header Access-Control-Allow-Credentials true;
+                add_header Access-Control-Max-Age 86400;
+                return 204;
+            }
             set $chat_upstream http://chat:3000;
             proxy_pass $chat_upstream;
             proxy_http_version 1.1;
@@ -104,9 +132,19 @@ http {
             proxy_set_header X-Forwarded-Proto $scheme;
             proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
             proxy_set_header Cookie $http_cookie;
+            add_header Access-Control-Allow-Origin $cors_origin always;
+            add_header Access-Control-Allow-Credentials true always;
         }
 
         location = /api/admin/session {
+            if ($request_method = OPTIONS) {
+                add_header Access-Control-Allow-Origin $cors_origin;
+                add_header Access-Control-Allow-Methods "GET, POST, PUT, DELETE, OPTIONS";
+                add_header Access-Control-Allow-Headers "Content-Type, Authorization";
+                add_header Access-Control-Allow-Credentials true;
+                add_header Access-Control-Max-Age 86400;
+                return 204;
+            }
             set $chat_upstream http://chat:3000;
             proxy_pass $chat_upstream;
             proxy_http_version 1.1;
@@ -115,9 +153,19 @@ http {
             proxy_set_header X-Forwarded-Proto $scheme;
             proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
             proxy_set_header Cookie $http_cookie;
+            add_header Access-Control-Allow-Origin $cors_origin always;
+            add_header Access-Control-Allow-Credentials true always;
         }
 
         location /api/admin/ {
+            if ($request_method = OPTIONS) {
+                add_header Access-Control-Allow-Origin $cors_origin;
+                add_header Access-Control-Allow-Methods "GET, POST, PUT, DELETE, OPTIONS";
+                add_header Access-Control-Allow-Headers "Content-Type, Authorization";
+                add_header Access-Control-Allow-Credentials true;
+                add_header Access-Control-Max-Age 86400;
+                return 204;
+            }
             auth_request /_admin_auth;
             error_page 401 = @admin_api_unauthorized;
             set $chat_upstream http://chat:3000;
@@ -128,9 +176,19 @@ http {
             proxy_set_header X-Forwarded-Proto $scheme;
             proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
             proxy_set_header Cookie $http_cookie;
+            add_header Access-Control-Allow-Origin $cors_origin always;
+            add_header Access-Control-Allow-Credentials true always;
         }
 
         location /api/ {
+            if ($request_method = OPTIONS) {
+                add_header Access-Control-Allow-Origin $cors_origin;
+                add_header Access-Control-Allow-Methods "GET, POST, PUT, DELETE, OPTIONS";
+                add_header Access-Control-Allow-Headers "Content-Type, Authorization";
+                add_header Access-Control-Allow-Credentials true;
+                add_header Access-Control-Max-Age 86400;
+                return 204;
+            }
             set $chat_upstream http://chat:3000;
             proxy_pass $chat_upstream;
             proxy_http_version 1.1;
@@ -138,6 +196,8 @@ http {
             proxy_set_header X-Forwarded-Host $host;
             proxy_set_header X-Forwarded-Proto $scheme;
             proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+            add_header Access-Control-Allow-Origin $cors_origin always;
+            add_header Access-Control-Allow-Credentials true always;
         }
 
         location = /metrics {

package.json

@@ -0,0 +1,5 @@
+{
+  "dependencies": {
+    "vitest": "^4.1.11"
+  }
+}

scripts/cache-hit-ratio.sh

@@ -0,0 +1,6 @@
+#!/usr/bin/env bash
+# cache-hit-ratio.sh — Thin wrapper that runs the Node.js cache-hit-ratio tool.
+# Mirrors scripts/cache-hit-ratio.ps1.
+set -euo pipefail
+script_dir="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+exec node "$script_dir/cache-hit-ratio.js" "$@"

scripts/check-env.ps1

@@ -0,0 +1,137 @@
+#!/usr/bin/env pwsh
+# check-env.ps1 — Verify the deployment environment for Linear Void.
+#
+# Checks, against the target Docker host (default test lab 192.168.88.40):
+#   1. Docker Engine reachability over the TCP socket (port 2375).
+#   2. Coolify helper stack (proxy/Traefik, sentinel, cloudflared).
+#   3. Linear Void containers (playout, chat, nginx, redis, db, prometheus, grafana).
+#   4. Public endpoints (app, HLS manifest, admin).
+#   5. Health endpoints exposed by the services.
+#
+# Usage:
+#   ./scripts/check-env.ps1
+#   ./scripts/check-env.ps1 -Target 192.168.88.40
+#   ./scripts/check-env.ps1 -Target 192.168.88.40 -DockerPort 2375 -OnlyContainers
+#
+# Exit code is 0 when all required checks pass, 1 otherwise.
+
+param(
+    [string]$Target = '192.168.88.40',
+    [int]$DockerPort = 2375,
+    [switch]$OnlyContainers,
+    [switch]$OnlyEndpoints,
+    [int]$TimeoutSec = 10
+)
+
+$ErrorActionPreference = 'Stop'
+$target = $Target
+$dockerHost = "tcp://${target}:${DockerPort}"
+
+$failed = $false
+function Fail([string]$msg) { Write-Host "  [FAIL] $msg" -ForegroundColor Red; $script:failed = $true }
+function Ok([string]$msg)   { Write-Host "  [OK]   $msg" -ForegroundColor Green }
+function Warn([string]$msg) { Write-Host "  [WARN] $msg" -ForegroundColor Yellow }
+
+Write-Host "== Linear Void environment check ==" -ForegroundColor Cyan
+Write-Host "Target Docker host : $dockerHost"
+Write-Host ("Date               : " + (Get-Date -Format s))
+
+# ---------------------------------------------------------------------------
+# 1. Docker Engine reachability
+# ---------------------------------------------------------------------------
+if (-not $OnlyEndpoints) {
+    Write-Host "`n-- Docker Engine (${target}:${DockerPort}) --" -ForegroundColor Cyan
+    $env:DOCKER_HOST = $dockerHost
+    try {
+        $ver = docker version --format '{{.Server.Version}}' 2>&1
+        if ($LASTEXITCODE -eq 0 -and $ver -match '\d') {
+            Ok "Docker Engine reachable (server version $ver)"
+        } else {
+            Fail "Docker Engine not reachable: $($ver -join ' ')"
+        }
+    } catch {
+        Fail "Docker Engine error: $($_.Exception.Message)"
+    }
+
+    # -----------------------------------------------------------------------
+    # 2. Coolify helper stack
+    # -----------------------------------------------------------------------
+    Write-Host "`n-- Coolify managed-server helpers --" -ForegroundColor Cyan
+    foreach ($c in @('coolify-proxy', 'coolify-sentinel', 'coolify-cloudflared')) {
+        try {
+            $state = docker inspect $c --format '{{.State.Status}} ({{.State.Health.Status}})' 2>&1
+            if ($LASTEXITCODE -eq 0) {
+                if ($state -match 'running.*(healthy|^running \(running\))') { Ok "coolify helper: $c -> $state" }
+                else { Warn "coolify helper: $c -> $state" }
+            } else {
+                Warn "coolify helper: $c -> NOT FOUND (is this a Coolify-managed server?)"
+            }
+        } catch { Warn "coolify helper: $c -> error $($_.Exception.Message)" }
+    }
+
+    # -----------------------------------------------------------------------
+    # 3. Linear Void containers
+    # -----------------------------------------------------------------------
+    Write-Host "`n-- Linear Void containers --" -ForegroundColor Cyan
+    $required = @('linearvoid-playout', 'linearvoid-chat', 'linearvoid-nginx',
+                  'linearvoid-redis', 'linearvoid-db', 'linearvoid-prometheus', 'linearvoid-grafana')
+    try {
+        $running = @(docker ps --format '{{.Names}}' 2>&1)
+        $all = @(docker ps -a --format '{{.Names}}\t{{.Status}}' 2>&1)
+        foreach ($name in $required) {
+            if ($running -contains $name) {
+                Ok "container: $name is running"
+            } elseif (($all | Where-Object { $_ -match "^$name\s" })) {
+                $status = ($all | Where-Object { $_ -match "^$name\s" } | Select-Object -First 1)
+                Fail "container: $name EXISTS but not running -> $status"
+            } else {
+                Fail "container: $name MISSING"
+            }
+        }
+    } catch { Fail "container listing error: $($_.Exception.Message)" }
+}
+
+# ---------------------------------------------------------------------------
+# 4. Health endpoints (via Docker host)
+# ---------------------------------------------------------------------------
+if (-not $OnlyContainers) {
+    Write-Host "`n-- Health endpoints --" -ForegroundColor Cyan
+    $checks = @(
+        @{ Name = 'chat /health';      Url = "http://${target}:3000/health" },
+        @{ Name = 'nginx /health';     Url = "http://${target}:8081/health" },
+        @{ Name = 'prometheus';        Url = "http://${target}:9090/-/healthy" },
+        @{ Name = 'grafana /api/health'; Url = "http://${target}:3001/api/health" }
+    )
+    foreach ($c in $checks) {
+        try {
+            $r = Invoke-WebRequest -Uri $c.Url -TimeoutSec $TimeoutSec -UseBasicParsing -ErrorAction Stop
+            if ($r.StatusCode -eq 200) { Ok "$($c.Name) -> 200" } else { Warn "$($c.Name) -> $($r.StatusCode)" }
+        } catch {
+            Fail "$($c.Name) -> $($_.Exception.Response.StatusCode.value__)"
+        }
+    }
+
+    # -----------------------------------------------------------------------
+    # 5. Public endpoints (routed through the Docker host for reliable TLS)
+    # -----------------------------------------------------------------------
+    Write-Host "`n-- Public endpoints (via Docker host) --" -ForegroundColor Cyan
+    $pub = @(
+        @{ Name = 'linearvoid.tv app';     Url = 'https://linearvoid.tv/' },
+        @{ Name = 'media HLS manifest';    Url = 'https://media.linearvoid.tv/hls/playlist.m3u8' },
+        @{ Name = 'coolify dashboard';     Url = 'https://coolify.allucanget.biz/' }
+    )
+    foreach ($c in $pub) {
+        try {
+            $code = (& docker run --rm curlimages/curl:latest -sS -o /dev/null -w '%{http_code}' --max-time $TimeoutSec -L $c.Url 2>&1 | Out-String).Trim()
+            if ($code -match '^\d{3}$') { Ok "$($c.Name) -> HTTP $code" }
+            else { Warn "$($c.Name) -> $code" }
+        } catch {
+            Warn "$($c.Name) -> error $($_.Exception.Message.Split("`n")[0])"
+        }
+    }
+}
+
+# ---------------------------------------------------------------------------
+Write-Host ""
+if ($failed) { Write-Host "RESULT: FAILED" -ForegroundColor Red; exit 1 }
+Write-Host "RESULT: OK" -ForegroundColor Green; exit 0

scripts/check-env.sh

@@ -0,0 +1,132 @@
+#!/usr/bin/env bash
+# check-env.sh — Verify the deployment environment for Linear Void.
+# Mirrors scripts/check-env.ps1.
+#
+# Checks, against the target Docker host (default test lab 192.168.88.40):
+#   1. Docker Engine reachability over the TCP socket (port 2375).
+#   2. Coolify helper stack (proxy/Traefik, sentinel, cloudflared).
+#   3. Linear Void containers (playout, chat, nginx, redis, db, prometheus, grafana).
+#   4. Public endpoints (app, HLS manifest, admin).
+#   5. Health endpoints exposed by the services.
+#
+# Usage:
+#   ./scripts/check-env.sh
+#   ./scripts/check-env.sh -t 192.168.88.40
+#   ./scripts/check-env.sh -t 192.168.88.40 -p 2375 --only-containers
+#
+# Exit code is 0 when all required checks pass, 1 otherwise.
+set -uo pipefail
+
+TARGET=192.168.88.40
+DOCKER_PORT=2375
+ONLY_CONTAINERS=0
+ONLY_ENDPOINTS=0
+TIMEOUT=10
+
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    -t|--target)   TARGET="$2"; shift 2;;
+    -p|--docker-port) DOCKER_PORT="$2"; shift 2;;
+    --only-containers) ONLY_CONTAINERS=1; shift;;
+    --only-endpoints)  ONLY_ENDPOINTS=1; shift;;
+    --timeout)     TIMEOUT="$2"; shift 2;;
+    *) echo "Unknown option: $1" >&2; exit 2;;
+  esac
+done
+
+export DOCKER_HOST="tcp://${TARGET}:${DOCKER_PORT}"
+
+FAILED=0
+fail() { printf '  [FAIL] %s\n' "$1"; FAILED=1; }
+ok()   { printf '  [OK]   %s\n' "$1"; }
+warn() { printf '  [WARN] %s\n' "$1"; }
+
+echo "== Linear Void environment check =="
+echo "Target Docker host : $DOCKER_HOST"
+echo "Date               : $(date -Iseconds)"
+
+# ---------------------------------------------------------------------------
+# 1. Docker Engine reachability
+# ---------------------------------------------------------------------------
+if [[ "$ONLY_ENDPOINTS" -eq 0 ]]; then
+  echo ""
+  echo "-- Docker Engine (${TARGET}:${DOCKER_PORT}) --"
+  if ver=$(docker version --format '{{.Server.Version}}' 2>&1) && [[ "$ver" =~ [0-9] ]]; then
+    ok "Docker Engine reachable (server version $ver)"
+  else
+    fail "Docker Engine not reachable: $ver"
+  fi
+
+  # -------------------------------------------------------------------------
+  # 2. Coolify helper stack
+  # -------------------------------------------------------------------------
+  echo ""
+  echo "-- Coolify managed-server helpers --"
+  for c in coolify-proxy coolify-sentinel coolify-cloudflared; do
+    if state=$(docker inspect "$c" --format '{{.State.Status}} ({{.State.Health.Status}})' 2>/dev/null); then
+      if [[ "$state" =~ running.*(healthy|running) ]]; then ok "coolify helper: $c -> $state"; else warn "coolify helper: $c -> $state"; fi
+    else
+      warn "coolify helper: $c -> NOT FOUND (is this a Coolify-managed server?)"
+    fi
+  done
+
+  # -------------------------------------------------------------------------
+  # 3. Linear Void containers
+  # -------------------------------------------------------------------------
+  echo ""
+  echo "-- Linear Void containers --"
+  required=(linearvoid-playout linearvoid-chat linearvoid-nginx linearvoid-redis linearvoid-db linearvoid-prometheus linearvoid-grafana)
+  running="$(docker ps --format '{{.Names}}' 2>/dev/null)"
+  all="$(docker ps -a --format '{{.Names}}\t{{.Status}}' 2>/dev/null)"
+  for name in "${required[@]}"; do
+    if grep -qx "$name" <<<"$running"; then
+      ok "container: $name is running"
+    elif grep -q "^$name[[:space:]]" <<<"$all"; then
+      status="$(grep "^$name[[:space:]]" <<<"$all" | head -1)"
+      fail "container: $name EXISTS but not running -> $status"
+    else
+      fail "container: $name MISSING"
+    fi
+  done
+fi
+
+# ---------------------------------------------------------------------------
+# 4. Health endpoints (via Docker host)
+# ---------------------------------------------------------------------------
+if [[ "$ONLY_CONTAINERS" -eq 0 ]]; then
+  echo ""
+  echo "-- Health endpoints --"
+  checks=(
+    "chat /health|http://${TARGET}:3000/health"
+    "nginx /health|http://${TARGET}:8081/health"
+    "prometheus|http://${TARGET}:9090/-/healthy"
+    "grafana /api/health|http://${TARGET}:3001/api/health"
+  )
+  for entry in "${checks[@]}"; do
+    name="${entry%%|*}"; url="${entry#*|}"
+    code=$(curl -s -o /dev/null -w '%{http_code}' --max-time "$TIMEOUT" "$url" 2>/dev/null)
+    if [[ "$code" == "200" ]]; then ok "$name -> 200"; elif [[ "$code" =~ ^[0-9]{3}$ ]]; then warn "$name -> $code"; else fail "$name -> unreachable"; fi
+  done
+
+  # -------------------------------------------------------------------------
+  # 5. Public endpoints (routed through the Docker host for reliable TLS)
+  # -------------------------------------------------------------------------
+  echo ""
+  echo "-- Public endpoints (via Docker host) --"
+  pub=(
+    "linearvoid.tv app|https://linearvoid.tv/"
+    "media HLS manifest|https://media.linearvoid.tv/hls/playlist.m3u8"
+    "coolify dashboard|https://coolify.allucanget.biz/"
+  )
+  for entry in "${pub[@]}"; do
+    name="${entry%%|*}"; url="${entry#*|}"
+    code=$(docker run --rm curlimages/curl:latest -sS -o /dev/null -w '%{http_code}' --max-time "$TIMEOUT" -L "$url" 2>/dev/null | tr -d '[:space:]')
+    if [[ "$code" =~ ^[0-9]{3}$ ]]; then ok "$name -> HTTP $code"; else warn "$name -> $code"; fi
+  done
+fi
+
+# ---------------------------------------------------------------------------
+echo ""
+if [[ "$FAILED" -ne 0 ]]; then echo "RESULT: FAILED"; exit 1; fi
+echo "RESULT: OK"
+exit 0

scripts/check-tokens.ps1

@@ -0,0 +1,176 @@
+#!/usr/bin/env pwsh
+# check-tokens.ps1 — Verify that the credentials in the environment actually work.
+#
+# Reads .env (or a supplied file) and validates access against live services:
+#   1. Gitea container registry (CI_REGISTRY_TOKEN / CI_REGISTRY_USER).
+#   2. Cloudflare API token(s) (CLOUDFLARE_API_TOKEN, and any R2 admin token).
+#   3. R2/S3 bucket access (R2_ACCESS_KEY_ID + R2_SECRET_ACCESS_KEY) — lists buckets.
+#   4. Linear Void admin login (ADMIN_USERNAME / ADMIN_PASSWORD) against the app origin.
+#
+# NOTE: Runs the checks through the target Docker host (default 192.168.88.40) so
+# network + TLS work regardless of this machine's proxy/firewall. Uses the
+# curlimages/curl and amazon/aws-cli images, which are pulled on demand.
+#
+# Usage:
+#   ./scripts/check-tokens.ps1
+#   ./scripts/check-tokens.ps1 -EnvFile .env
+#   ./scripts/check-tokens.ps1 -Target 192.168.88.40
+#
+# Exit code 0 = all credentials valid; 1 = at least one check failed.
+
+param(
+    [string]$EnvFile = '',
+    [string]$Target = '192.168.88.40',
+    [int]$DockerPort = 2375,
+    [int]$TimeoutSec = 20
+)
+
+$ErrorActionPreference = 'Stop'
+$target = $Target
+$dockerHost = "tcp://${target}:${DockerPort}"
+$env:DOCKER_HOST = $dockerHost
+
+$failed = $false
+function Fail([string]$msg) { Write-Host "  [FAIL] $msg" -ForegroundColor Red; $script:failed = $true }
+function Ok([string]$msg)   { Write-Host "  [OK]   $msg" -ForegroundColor Green }
+function Warn([string]$msg) { Write-Host "  [WARN] $msg" -ForegroundColor Yellow }
+
+# Resolve env file: explicit, or $PWD/.env, or repo root.
+if (-not $EnvFile) {
+    foreach ($cand in @((Join-Path $PWD '.env'), (Join-Path (Split-Path $PWD) '.env'))) {
+        if (Test-Path $cand) { $EnvFile = $cand; break }
+    }
+}
+if (-not $EnvFile -or -not (Test-Path $EnvFile)) {
+    Write-Host "No .env file found. Pass one with -EnvFile." -ForegroundColor Yellow
+    exit 1
+}
+
+# Parse KEY=VALUE lines into a hashtable.
+$envs = @{}
+Get-Content $EnvFile | ForEach-Object {
+    if ($_ -match '^\s*([A-Za-z0-9_]+)\s*=\s*(.*?)\s*$') {
+        $envs[$Matches[1]] = $Matches[2]
+    }
+}
+
+function Get-Val([string]$key) {
+    if ($envs.ContainsKey($key)) { return $envs[$key] }
+    return $null
+}
+
+# Pull helper images once.
+Write-Host "Preparing helper images (curl, aws-cli) on $target..." -ForegroundColor Cyan
+docker pull curlimages/curl:latest 2>&1 | Out-Null
+docker pull amazon/aws-cli:latest 2>&1 | Out-Null
+
+function Curl([string]$url, [string]$bearer, [string]$tokenHeader) {
+    $a = @('run', '--rm', 'curlimages/curl:latest', '-sS', '--max-time', "$TimeoutSec")
+    if ($bearer) { $a += @('-H', "Authorization: Bearer $bearer") }
+    if ($tokenHeader) { $a += @('-H', "Authorization: token $tokenHeader") }
+    $a += $url
+    return (& docker @a 2>&1 | Out-String).Trim()
+}
+
+# ---------------------------------------------------------------------------
+Write-Host "== Linear Void credential / token check ==" -ForegroundColor Cyan
+Write-Host "Env file   : $EnvFile"
+Write-Host "Docker host: $dockerHost"
+
+# --- 1. Gitea registry token ---
+Write-Host "`n-- Gitea container registry --" -ForegroundColor Cyan
+$regUser = Get-Val 'CI_REGISTRY_USER'
+$regToken = Get-Val 'CI_REGISTRY_TOKEN'
+if ($regUser -and $regToken) {
+    # Non-sensitive probe: /api/v1/user returns a scope message for token-based creds.
+    $out = Curl 'https://git.allucanget.biz/api/v1/user' $null $regToken
+    if ($out -match 'token does not have at least one of required scope' -or $out -match '"login"' -or $out -match 'invalid_grant') {
+        if ($out -match 'invalid_grant|Unauthorized|invalid') {
+            Fail "Gitea registry token rejected ($regUser): $out"
+        } else {
+            Ok "Gitea registry token valid for '$regUser' (scope: write:package/write:repository)"
+        }
+    } else {
+        Warn "Gitea registry token response unexpected: $out"
+    }
+} else {
+    Warn "CI_REGISTRY_USER / CI_REGISTRY_TOKEN not set; skipping Gitea registry check"
+}
+
+# --- 2. Cloudflare API token ---
+Write-Host "`n-- Cloudflare API tokens --" -ForegroundColor Cyan
+$cfToken = Get-Val 'CLOUDFLARE_API_TOKEN'
+if ($cfToken) {
+    $out = Curl 'https://api.cloudflare.com/client/v4/user/tokens/verify' $cfToken $null
+    if ($out -match '"success":true') {
+        if ($out -match '"status":"active"') { Ok 'Cloudflare API token valid and ACTIVE' }
+        else { Warn "Cloudflare API token valid but status: $out" }
+    } else {
+        Fail "Cloudflare API token rejected: $out"
+    }
+} else { Warn "CLOUDFLARE_API_TOKEN not set; skipping Cloudflare check" }
+
+# --- 3. Cloudflare R2 / S3 bucket access ---
+Write-Host "`n-- Cloudflare R2 (S3-compatible) --" -ForegroundColor Cyan
+$r2Key = Get-Val 'R2_ACCESS_KEY_ID'
+$r2Secret = Get-Val 'R2_SECRET_ACCESS_KEY'
+$r2Ep = Get-Val 'R2_ENDPOINT'
+$bucket = Get-Val 'R2_BUCKET_NAME'
+
+# Use the admin key pair (which has ListBuckets) when present, else the regular key.
+$listKey = if ((Get-Val 'R2_ADMIN_ACCESS_KEY_ID') -and (Get-Val 'R2_ADMIN_SECRET_ACCESS_KEY')) {
+    @{ k = Get-Val 'R2_ADMIN_ACCESS_KEY_ID'; s = Get-Val 'R2_ADMIN_SECRET_ACCESS_KEY' }
+} else { @{ k = $r2Key; s = $r2Secret } }
+
+if ($listKey.k -and $listKey.s -and $r2Ep) {
+    $script = @"
+export AWS_ACCESS_KEY_ID='$($listKey.k)'
+export AWS_SECRET_ACCESS_KEY='$($listKey.s)'
+aws s3 ls --endpoint-url '$r2Ep' --region auto 2>&1
+"@
+    $out = (& docker run --rm --entrypoint sh amazon/aws-cli:latest -c $script 2>&1 | Out-String).Trim()
+    if ($out -match $bucket) { Ok "R2 bucket listing OK, '$bucket' present" }
+    elseif ($out -match 'AccessDenied') { Fail "R2 listing denied (no ListBuckets permission): $out" }
+    elseif ($out -match '^\S+\s+') { Warn "R2 listing returned (unexpected): $out" }
+    else { Fail "R2 access failed: $out" }
+} else { Warn "R2 credentials / endpoint not fully set; skipping R2 check" }
+
+# Verify the specific application bucket is reachable with the primary key (read).
+if ($r2Key -and $r2Secret -and $r2Ep -and $bucket) {
+    $script = @"
+export AWS_ACCESS_KEY_ID='$r2Key'
+export AWS_SECRET_ACCESS_KEY='$r2Secret'
+aws s3 ls "s3://$bucket/" --endpoint-url '$r2Ep' --region auto --max-keys 5 2>&1 | head -5
+"@
+    $out = (& docker run --rm --entrypoint sh amazon/aws-cli:latest -c $script 2>&1 | Out-String).Trim()
+    if ($out -match 'AccessDenied') { Fail "Bucket '$bucket' read denied with primary key" }
+    elseif ($out -eq '') { Ok "Bucket '$bucket' reachable (empty listing, expected)" }
+    elseif ($out -match 'NoSuchBucket') { Fail "Bucket '$bucket' NOT FOUND" }
+    else { Ok "Bucket '$bucket' reachable" }
+} elseif (-not $bucket) { Warn "R2_BUCKET_NAME not set; skipping bucket-level check" }
+
+# --- 4. Linear Void admin login ---
+Write-Host "`n-- Linear Void admin login --" -ForegroundColor Cyan
+$adminUser = Get-Val 'ADMIN_USERNAME'
+$adminPass = Get-Val 'ADMIN_PASSWORD'
+if ($adminUser -and $adminPass) {
+    # Hit the chat backend directly (public origin is caught by nginx SPA fallback).
+    $url = "http://${target}:3000/api/admin/login"
+    $body = ('{"username":"' + $adminUser + '","password":"' + $adminPass + '"}')
+    $wrongBody = ('{"username":"' + $adminUser + '","password":"__definitely_wrong__"}')
+    $a = @('run', '--rm', 'curlimages/curl:latest', '-sS', '--max-time', "$TimeoutSec",
+           '-o', '/dev/null', '-w', '%{http_code}', '-X', 'POST',
+           '-H', 'Content-Type: application/json', '-d', $body, $url)
+    $okCode = (& docker @a 2>&1 | Out-String).Trim()
+    $b = @('run', '--rm', 'curlimages/curl:latest', '-sS', '--max-time', "$TimeoutSec",
+           '-o', '/dev/null', '-w', '%{http_code}', '-X', 'POST',
+           '-H', 'Content-Type: application/json', '-d', $wrongBody, $url)
+    $wrongCode = (& docker @b 2>&1 | Out-String).Trim()
+    if ($okCode -eq '200' -and $wrongCode -eq '401') { Ok "Admin credentials valid (correct=200, wrong=401)" }
+    else { Fail "Admin login unexpected (correct=$okCode, wrong=$wrongCode)" }
+} else { Warn "ADMIN_USERNAME / ADMIN_PASSWORD not set; skipping admin check" }
+
+# ---------------------------------------------------------------------------
+Write-Host ""
+if ($failed) { Write-Host "RESULT: FAILED" -ForegroundColor Red; exit 1 }
+Write-Host "RESULT: OK (no failing checks)" -ForegroundColor Green; exit 0

scripts/check-tokens.sh

@@ -0,0 +1,177 @@
+#!/usr/bin/env bash
+# check-tokens.sh — Verify that the credentials in the environment actually work.
+# Mirrors scripts/check-tokens.ps1.
+#
+# Reads .env (or a supplied file) and validates access against live services:
+#   1. Gitea container registry (CI_REGISTRY_TOKEN / CI_REGISTRY_USER).
+#   2. Cloudflare API token(s) (CLOUDFLARE_API_TOKEN, and any R2 admin token).
+#   3. R2/S3 bucket access (R2_ACCESS_KEY_ID + R2_SECRET_ACCESS_KEY) — lists buckets.
+#   4. Linear Void admin login (ADMIN_USERNAME / ADMIN_PASSWORD) against the app origin.
+#
+# NOTE: Runs the checks through the target Docker host (default 192.168.88.40) so
+# network + TLS work regardless of this machine's proxy/firewall. Uses the
+# curlimages/curl and amazon/aws-cli images, which are pulled on demand.
+#
+# Usage:
+#   ./scripts/check-tokens.sh
+#   ./scripts/check-tokens.sh -e .env
+#   ./scripts/check-tokens.sh -t 192.168.88.40
+#
+# Exit code 0 = all credentials valid; 1 = at least one check failed.
+set -uo pipefail
+
+ENV_FILE=""
+TARGET=192.168.88.40
+DOCKER_PORT=2375
+TIMEOUT=20
+
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    -e|--env-file) ENV_FILE="$2"; shift 2;;
+    -t|--target)   TARGET="$2"; shift 2;;
+    -p|--docker-port) DOCKER_PORT="$2"; shift 2;;
+    --timeout)     TIMEOUT="$2"; shift 2;;
+    *) echo "Unknown option: $1" >&2; exit 2;;
+  esac
+done
+
+export DOCKER_HOST="tcp://${TARGET}:${DOCKER_PORT}"
+
+FAILED=0
+fail() { printf '  [FAIL] %s\n' "$1"; FAILED=1; }
+ok()   { printf '  [OK]   %s\n' "$1"; }
+warn() { printf '  [WARN] %s\n' "$1"; }
+
+# Resolve env file: explicit, or $PWD/.env, or repo root.
+if [[ -z "$ENV_FILE" ]]; then
+  for cand in "$PWD/.env" "$(dirname "$PWD")/.env"; do
+    if [[ -f "$cand" ]]; then ENV_FILE="$cand"; break; fi
+  done
+fi
+if [[ -z "$ENV_FILE" || ! -f "$ENV_FILE" ]]; then
+  echo "No .env file found. Pass one with -e." >&2
+  exit 1
+fi
+
+# Parse KEY=VALUE lines into an associative array.
+declare -A envs
+while IFS= read -r line || [[ -n "$line" ]]; do
+  if [[ "$line" =~ ^[[:space:]]*([A-Za-z0-9_]+)[[:space:]]*=[[:space:]]*(.*)[[:space:]]*$ ]]; then
+    envs["${BASH_REMATCH[1]}"]="${BASH_REMATCH[2]}"
+  fi
+done < "$ENV_FILE"
+
+get_val() { [[ -n "${envs[$1]+x}" ]] && echo "${envs[$1]}"; }
+
+# Pull helper images once.
+echo "Preparing helper images (curl, aws-cli) on $TARGET..."
+docker pull curlimages/curl:latest >/dev/null 2>&1 || true
+docker pull amazon/aws-cli:latest >/dev/null 2>&1 || true
+
+# dc_curl URL [bearer] [token] — run curl via the curlimages/curl container.
+dc_curl() {
+  local url="$1" bearer="${2:-}" token="${3:-}"
+  local -a args=(run --rm curlimages/curl:latest -sS --max-time "$TIMEOUT")
+  [[ -n "$bearer" ]] && args+=( -H "Authorization: Bearer $bearer" )
+  [[ -n "$token" ]] && args+=( -H "Authorization: token $token" )
+  args+=( "$url" )
+  docker "${args[@]}" 2>&1 | tr -d '\r' | sed -e 's/^[[:space:]]*//' -e 's/[[:space:]]*$//'
+}
+
+# ---------------------------------------------------------------------------
+echo "== Linear Void credential / token check =="
+echo "Env file   : $ENV_FILE"
+echo "Docker host: $DOCKER_HOST"
+
+# --- 1. Gitea registry token ---
+echo ""
+echo "-- Gitea container registry --"
+reg_user="$(get_val CI_REGISTRY_USER)"
+reg_token="$(get_val CI_REGISTRY_TOKEN)"
+if [[ -n "$reg_user" && -n "$reg_token" ]]; then
+  # Non-sensitive probe: /api/v1/user returns a scope message for token-based creds.
+  out="$(dc_curl 'https://git.allucanget.biz/api/v1/user' "" "$reg_token")"
+  if [[ "$out" =~ token.does.not.have.at.least.one.of.required.scope || "$out" =~ '"login"' || "$out" =~ invalid_grant ]]; then
+    if [[ "$out" =~ invalid_grant|Unauthorized|invalid ]]; then
+      fail "Gitea registry token rejected ($reg_user): $out"
+    else
+      ok "Gitea registry token valid for '$reg_user' (scope: write:package/write:repository)"
+    fi
+  else
+    warn "Gitea registry token response unexpected: $out"
+  fi
+else
+  warn "CI_REGISTRY_USER / CI_REGISTRY_TOKEN not set; skipping Gitea registry check"
+fi
+
+# --- 2. Cloudflare API token ---
+echo ""
+echo "-- Cloudflare API tokens --"
+cf_token="$(get_val CLOUDFLARE_API_TOKEN)"
+if [[ -n "$cf_token" ]]; then
+  out="$(dc_curl 'https://api.cloudflare.com/client/v4/user/tokens/verify' "$cf_token")"
+  if [[ "$out" =~ '"success":true' ]]; then
+    if [[ "$out" =~ '"status":"active"' ]]; then ok 'Cloudflare API token valid and ACTIVE'; else warn "Cloudflare API token valid but status: $out"; fi
+  else
+    fail "Cloudflare API token rejected: $out"
+  fi
+else
+  warn "CLOUDFLARE_API_TOKEN not set; skipping Cloudflare check"
+fi
+
+# --- 3. Cloudflare R2 / S3 bucket access ---
+echo ""
+echo "-- Cloudflare R2 (S3-compatible) --"
+r2_key="$(get_val R2_ACCESS_KEY_ID)"
+r2_secret="$(get_val R2_SECRET_ACCESS_KEY)"
+r2_ep="$(get_val R2_ENDPOINT)"
+bucket="$(get_val R2_BUCKET_NAME)"
+
+# Use the admin key pair (which has ListBuckets) when present, else the regular key.
+list_k="$(get_val R2_ADMIN_ACCESS_KEY_ID)"; list_s="$(get_val R2_ADMIN_SECRET_ACCESS_KEY)"
+if [[ -z "$list_k" || -z "$list_s" ]]; then list_k="$r2_key"; list_s="$r2_secret"; fi
+
+if [[ -n "$list_k" && -n "$list_s" && -n "$r2_ep" ]]; then
+  out="$(docker run --rm --entrypoint sh amazon/aws-cli:latest -c "export AWS_ACCESS_KEY_ID='$list_k'; export AWS_SECRET_ACCESS_KEY='$list_s'; aws s3 ls --endpoint-url '$r2_ep' --region auto 2>&1" 2>&1)"
+  if grep -q "$bucket" <<<"$out"; then ok "R2 bucket listing OK, '$bucket' present"
+  elif grep -q 'AccessDenied' <<<"$out"; then fail "R2 listing denied (no ListBuckets permission): $out"
+  elif grep -qE '^[^[:space:]]+[[:space:]]+' <<<"$out"; then warn "R2 listing returned (unexpected): $out"
+  else fail "R2 access failed: $out"; fi
+else
+  warn "R2 credentials / endpoint not fully set; skipping R2 check"
+fi
+
+# Verify the specific application bucket is reachable with the primary key (read).
+if [[ -n "$r2_key" && -n "$r2_secret" && -n "$r2_ep" && -n "$bucket" ]]; then
+  out="$(docker run --rm --entrypoint sh amazon/aws-cli:latest -c "export AWS_ACCESS_KEY_ID='$r2_key'; export AWS_SECRET_ACCESS_KEY='$r2_secret'; aws s3 ls s3://$bucket/ --endpoint-url '$r2_ep' --region auto --max-keys 5 2>&1 | head -5" 2>&1)"
+  if grep -q 'AccessDenied' <<<"$out"; then fail "Bucket '$bucket' read denied with primary key"
+  elif [[ -z "$out" ]]; then ok "Bucket '$bucket' reachable (empty listing, expected)"
+  elif grep -q 'NoSuchBucket' <<<"$out"; then fail "Bucket '$bucket' NOT FOUND"
+  else ok "Bucket '$bucket' reachable"; fi
+elif [[ -z "$bucket" ]]; then
+  warn "R2_BUCKET_NAME not set; skipping bucket-level check"
+fi
+
+# --- 4. Linear Void admin login ---
+echo ""
+echo "-- Linear Void admin login --"
+admin_user="$(get_val ADMIN_USERNAME)"
+admin_pass="$(get_val ADMIN_PASSWORD)"
+if [[ -n "$admin_user" && -n "$admin_pass" ]]; then
+  # Hit the chat backend directly (public origin is caught by nginx SPA fallback).
+  url="http://${TARGET}:3000/api/admin/login"
+  body="{\"username\":\"$admin_user\",\"password\":\"$admin_pass\"}"
+  wrong_body="{\"username\":\"$admin_user\",\"password\":\"__definitely_wrong__\"}"
+  ok_code="$(docker run --rm curlimages/curl:latest -sS --max-time "$TIMEOUT" -o /dev/null -w '%{http_code}' -X POST -H 'Content-Type: application/json' -d "$body" "$url" 2>/dev/null | tr -d '[:space:]')"
+  wrong_code="$(docker run --rm curlimages/curl:latest -sS --max-time "$TIMEOUT" -o /dev/null -w '%{http_code}' -X POST -H 'Content-Type: application/json' -d "$wrong_body" "$url" 2>/dev/null | tr -d '[:space:]')"
+  if [[ "$ok_code" == "200" && "$wrong_code" == "401" ]]; then ok "Admin credentials valid (correct=200, wrong=401)"
+  else fail "Admin login unexpected (correct=$ok_code, wrong=$wrong_code)"; fi
+else
+  warn "ADMIN_USERNAME / ADMIN_PASSWORD not set; skipping admin check"
+fi
+
+# ---------------------------------------------------------------------------
+echo ""
+if [[ "$FAILED" -ne 0 ]]; then echo "RESULT: FAILED"; exit 1; fi
+echo "RESULT: OK (no failing checks)"
+exit 0

scripts/loadtest/README.md

@@ -2,66 +2,42 @@
 
 These files back the [scaling/performance playbook](../../docs/operations/scaling-performance-playbook.md).
 
-## WebSocket chat
+## WebSocket chat (artillery)
 
-Run with an explicit WebSocket URL, for example:
+Run with an explicit WebSocket URL, capturing output to an artifact file:
 
 ```bash
-WS_URL=wss://linearvoid.tv/ws?username=loadtest&room=main npx artillery run scripts/loadtest/ws-chat.yml
-```
-
-Capture output to an artifact file for the worksheet:
-
-```bash
-WS_URL=wss://linearvoid.tv/ws?username=loadtest&room=main npx artillery run scripts/loadtest/ws-chat.yml --output artifacts/ws-chat-report.json
-```
-
-PowerShell equivalent:
-
-```powershell
-$env:WS_URL = 'wss://linearvoid.tv/ws?username=loadtest&room=main'
-New-Item -ItemType Directory -Force artifacts | Out-Null
+export WS_URL='wss://linearvoid.tv/ws?username=loadtest&room=main'
+mkdir -p artifacts
 npx artillery run scripts/loadtest/ws-chat.yml --output artifacts/ws-chat-report.json
 ```
 
 Summarize key counters, rates, and available percentile stats from a saved Artillery JSON report:
 
-```powershell
-pwsh -File scripts/loadtest/summarize-artillery.ps1 -ReportFile artifacts/ws-chat-report.json
+```bash
+scripts/loadtest/summarize-artillery.sh artifacts/ws-chat-report.json
 ```
 
-## Admin API
+## Admin API (k6)
 
-Run with a cookie or session header that can access admin endpoints:
+Run with a cookie or session header that can access admin endpoints, capturing machine-readable metrics and a text summary:
 
 ```bash
-ADMIN_BASE_URL=https://linearvoid.tv ADMIN_SESSION_COOKIE="session=..." k6 run scripts/loadtest/admin-api.js
-```
-
-Capture machine-readable metrics and a text summary:
-
-```bash
-ADMIN_BASE_URL=https://linearvoid.tv ADMIN_SESSION_COOKIE="session=..." k6 run --summary-export artifacts/admin-api-summary.json scripts/loadtest/admin-api.js > artifacts/admin-api-output.txt
-```
-
-PowerShell equivalent:
-
-```powershell
-$env:ADMIN_BASE_URL = 'https://linearvoid.tv'
-$env:ADMIN_SESSION_COOKIE = 'session=...'
-New-Item -ItemType Directory -Force artifacts | Out-Null
-k6 run --summary-export artifacts/admin-api-summary.json scripts/loadtest/admin-api.js | Tee-Object -FilePath artifacts/admin-api-output.txt
+export ADMIN_BASE_URL='https://linearvoid.tv'
+export ADMIN_SESSION_COOKIE='session=...'
+mkdir -p artifacts
+k6 run --summary-export artifacts/admin-api-summary.json scripts/loadtest/admin-api.js | tee artifacts/admin-api-output.txt
 ```
 
 The script emits global and per-endpoint thresholds (overview, analytics, chat-log, users) so p95 and error-rate regressions are visible per route.
 
 Summarize endpoint p95 and failure rates from a saved k6 JSON summary:
 
-```powershell
-pwsh -File scripts/loadtest/summarize-k6.ps1 -SummaryFile artifacts/admin-api-summary.json
+```bash
+scripts/loadtest/summarize-k6.sh artifacts/admin-api-summary.json
 ```
 
-## PowerShell helper
+## Runner helper
 
 Use the helper to run one scenario or both and automatically write outputs into a dated artifact folder:
 
@@ -69,16 +45,11 @@ Use the helper to run one scenario or both and automatically write outputs into
 - WebSocket runs also emit `ws-chat-summary.txt` with connection/error counters and available summary percentiles.
 - Admin runs also emit `admin-api-endpoint-summary.txt` with per-endpoint p95/error values.
 
-```powershell
-pwsh -File scripts/loadtest/run-loadtests.ps1 -Scenario ws -WsUrl 'wss://linearvoid.tv/ws?username=loadtest&room=main'
-```
-
-```powershell
-pwsh -File scripts/loadtest/run-loadtests.ps1 -Scenario admin -AdminBaseUrl 'https://linearvoid.tv' -AdminSessionCookie 'session=...'
-```
-
-```powershell
-pwsh -File scripts/loadtest/run-loadtests.ps1 -Scenario all -WsUrl 'wss://linearvoid.tv/ws?username=loadtest&room=main' -AdminBaseUrl 'https://linearvoid.tv' -AdminSessionCookie 'session=...'
+```bash
+scripts/loadtest/run-loadtests.sh -s ws -w 'wss://linearvoid.tv/ws?username=loadtest&room=main'
+scripts/loadtest/run-loadtests.sh -s admin -a 'https://linearvoid.tv' -c 'session=...'
+scripts/loadtest/run-loadtests.sh -s all -w 'wss://linearvoid.tv/ws?username=loadtest&room=main' \
+  -a 'https://linearvoid.tv' -c 'session=...'
 ```
 
 ## Notes

scripts/loadtest/run-loadtests.sh

@@ -0,0 +1,72 @@
+#!/usr/bin/env bash
+# run-loadtests.sh — Run an artillery (ws) and/or k6 (admin) load-test scenario
+# and write outputs into a dated artifacts folder.
+# Mirrors scripts/loadtest/run-loadtests.ps1.
+#
+# Usage:
+#   ./scripts/loadtest/run-loadtests.sh -s ws   -w 'wss://...'
+#   ./scripts/loadtest/run-loadtests.sh -s admin -a 'https://...' -c 'session=...'
+#   ./scripts/loadtest/run-loadtests.sh -s all  -w 'wss://...' -a 'https://...' -c 'session=...'
+set -euo pipefail
+
+script_dir="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+repo_root="$(cd "$script_dir/../.." && pwd)"
+
+SCENARIO="all"
+ARTIFACT_ROOT="artifacts"
+WS_URL="${WS_URL:-}"
+ADMIN_BASE_URL="${ADMIN_BASE_URL:-}"
+ADMIN_SESSION_COOKIE="${ADMIN_SESSION_COOKIE:-}"
+
+usage() {
+  echo "Usage: $0 [-s ws|admin|all] [-w ws_url] [-a admin_base_url] [-c admin_session_cookie] [-r artifact_root]" >&2
+  exit 2
+}
+
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    -s) SCENARIO="$2"; shift 2;;
+    -w) WS_URL="$2"; shift 2;;
+    -a) ADMIN_BASE_URL="$2"; shift 2;;
+    -c) ADMIN_SESSION_COOKIE="$2"; shift 2;;
+    -r) ARTIFACT_ROOT="$2"; shift 2;;
+    -h|--help) usage;;
+    *) echo "Unknown option: $1" >&2; usage;;
+  esac
+done
+
+case "$SCENARIO" in ws|admin|all) ;; *) echo "Scenario must be ws|admin|all" >&2; exit 2;; esac
+
+if [[ "$SCENARIO" == ws || "$SCENARIO" == all ]] && ! command -v npx >/dev/null 2>&1; then
+  echo "npx is required for ws scenario but was not found in PATH." >&2; exit 1
+fi
+if [[ "$SCENARIO" == admin || "$SCENARIO" == all ]] && ! command -v k6 >/dev/null 2>&1; then
+  echo "k6 is required for admin scenario but was not found in PATH." >&2; exit 1
+fi
+
+if [[ "$ARTIFACT_ROOT" = /* ]]; then artifact_base="$ARTIFACT_ROOT"; else artifact_base="$repo_root/$ARTIFACT_ROOT"; fi
+artifact_dir="$artifact_base/$(date +%Y-%m-%d_%H%M%S)"
+mkdir -p "$artifact_dir"
+
+if [[ "$SCENARIO" == ws || "$SCENARIO" == all ]]; then
+  [[ -n "$WS_URL" ]] || { echo "WS_URL is required for ws scenario. Pass -w or set \$WS_URL." >&2; exit 1; }
+  ws_out="$artifact_dir/ws-chat-report.json"
+  ws_sum="$artifact_dir/ws-chat-summary.txt"
+  echo "== Running artillery ws scenario =="
+  WS_URL="$WS_URL" npx artillery run "$script_dir/ws-chat.yml" --output "$ws_out"
+  "$script_dir/summarize-artillery.sh" "$ws_out" | tee "$ws_sum"
+fi
+
+if [[ "$SCENARIO" == admin || "$SCENARIO" == all ]]; then
+  [[ -n "$ADMIN_BASE_URL" ]] || { echo "ADMIN_BASE_URL is required for admin scenario. Pass -a or set \$ADMIN_BASE_URL." >&2; exit 1; }
+  [[ -n "$ADMIN_SESSION_COOKIE" ]] || { echo "ADMIN_SESSION_COOKIE is required for admin scenario. Pass -c or set \$ADMIN_SESSION_COOKIE." >&2; exit 1; }
+  admin_sum="$artifact_dir/admin-api-summary.json"
+  admin_out="$artifact_dir/admin-api-output.txt"
+  admin_ep_sum="$artifact_dir/admin-api-endpoint-summary.txt"
+  echo "== Running k6 admin scenario =="
+  ADMIN_BASE_URL="$ADMIN_BASE_URL" ADMIN_SESSION_COOKIE="$ADMIN_SESSION_COOKIE" \
+    k6 run --summary-export "$admin_sum" "$script_dir/admin-api.js" | tee "$admin_out"
+  "$script_dir/summarize-k6.sh" "$admin_sum" | tee "$admin_ep_sum"
+fi
+
+echo "Artifacts written to $artifact_dir"

scripts/loadtest/summarize-artillery.sh

@@ -0,0 +1,54 @@
+#!/usr/bin/env bash
+# summarize-artillery.sh — Extract counters, rates and percentile stats from an
+# Artillery JSON report. Mirrors scripts/loadtest/summarize-artillery.ps1.
+# Usage: ./scripts/loadtest/summarize-artillery.sh <artillery-report.json>
+set -euo pipefail
+
+if [[ $# -lt 1 ]]; then
+  echo "Usage: $0 <artillery-report.json>" >&2
+  exit 2
+fi
+report_file="$1"
+[[ -f "$report_file" ]] || { echo "Report file not found: $report_file" >&2; exit 1; }
+
+REPORT_FILE="$report_file" node <<'EOF'
+const fs = require('fs');
+const report = JSON.parse(fs.readFileSync(process.env.REPORT_FILE, 'utf8'));
+const aggregate = report.aggregate;
+if (!aggregate) throw new Error('Artillery report is missing aggregate data.');
+
+const get = (obj, name) => (obj && obj[name] !== undefined) ? obj[name] : undefined;
+const counters = aggregate.counters || {};
+const rates = aggregate.rates || {};
+const summaries = aggregate.summaries || {};
+
+const created = Number(get(counters, 'vusers.created') || 0);
+const failed = Number(get(counters, 'vusers.failed') || 0);
+const completed = Number(get(counters, 'vusers.completed') || 0);
+const errorRate = created > 0 ? +((failed / created) * 100).toFixed(2) : 0;
+
+console.log('=== Artillery WebSocket Summary ===');
+console.log(`source: ${process.env.REPORT_FILE}`);
+if (report.timestamp) console.log(`timestamp: ${report.timestamp}`);
+console.log(`vusers.created: ${created}`);
+console.log(`vusers.completed: ${completed}`);
+console.log(`vusers.failed: ${failed}`);
+console.log(`vusers.failure_rate_percent: ${errorRate}`);
+
+for (const metric of ['ws.messages_sent', 'ws.messages_received', 'ws.ping', 'plugins.metrics-by-endpoint.response_time']) {
+  const v = get(counters, metric);
+  if (v !== undefined) console.log(`${metric}: ${v}`);
+}
+for (const [k, v] of Object.entries(rates || {})) console.log(`rate.${k}: ${v}`);
+
+const fmt = (prefix, stats) => {
+  if (!stats) return;
+  const parts = [];
+  for (const name of ['count', 'min', 'max', 'mean', 'median', 'p95', 'p99']) {
+    const v = get(stats, name);
+    if (v !== undefined) parts.push(`${name}=${v}`);
+  }
+  if (parts.length) console.log(`${prefix}: ${parts.join(', ')}`);
+};
+for (const [k, v] of Object.entries(summaries)) fmt(`summary.${k}`, v);
+EOF

scripts/loadtest/summarize-k6.sh

@@ -0,0 +1,43 @@
+#!/usr/bin/env bash
+# summarize-k6.sh — Extract per-endpoint p95 and failure rates from a k6 JSON
+# summary export. Mirrors scripts/loadtest/summarize-k6.ps1.
+# Usage: ./scripts/loadtest/summarize-k6.sh <k6-summary.json>
+set -euo pipefail
+
+if [[ $# -lt 1 ]]; then
+  echo "Usage: $0 <k6-summary.json>" >&2
+  exit 2
+fi
+summary_file="$1"
+[[ -f "$summary_file" ]] || { echo "Summary file not found: $summary_file" >&2; exit 1; }
+
+SUMMARY_FILE="$summary_file" node <<'EOF'
+const fs = require('fs');
+const summary = JSON.parse(fs.readFileSync(process.env.SUMMARY_FILE, 'utf8'));
+const metrics = summary.metrics;
+if (!metrics) throw new Error('Invalid k6 summary file: missing metrics object.');
+
+const val = (m, key) => (m && m.values && m.values[key] !== undefined) ? Number(m.values[key]) : undefined;
+const globalDuration = val(metrics.http_req_duration, 'p(95)');
+const globalFailRate = val(metrics.http_req_failed, 'rate');
+
+console.log(`k6 Summary: ${process.env.SUMMARY_FILE}`);
+if (globalDuration !== undefined) console.log(`Global p95 duration (ms): ${globalDuration.toFixed(2)}`);
+if (globalFailRate !== undefined) console.log(`Global failure rate: ${(globalFailRate * 100).toFixed(2)}%`);
+console.log('');
+
+const endpoints = ['overview', 'analytics', 'chat_log', 'users'];
+const rows = endpoints.map((ep) => {
+  const dP95 = val(metrics[`http_req_duration{endpoint:${ep}}`], 'p(95)');
+  const fRate = val(metrics[`http_req_failed{endpoint:${ep}}`], 'rate');
+  return {
+    Endpoint: ep,
+    P95Ms: dP95 !== undefined ? dP95.toFixed(2) : 'n/a',
+    FailurePct: fRate !== undefined ? (fRate * 100).toFixed(4) : 'n/a',
+  };
+});
+
+const pad = (s, n) => String(s).padEnd(n);
+console.log(pad('ENDPOINT', 12) + pad('P95_MS', 10) + 'FAILURE_PCT');
+for (const r of rows) console.log(pad(r.Endpoint, 12) + pad(r.P95Ms, 10) + r.FailurePct);
+EOF