Add openrouter_judge.py to rank live image+JSON models by advertised price,
with ordered fallbacks, auth-error handling and auto selection. Wire both judges
to --model auto, record the model used, and add fake-catalogue/mocked-call
tests. Re-run the judges with the valid key: 13/13 screen designs match, 0 diff;
the Forward+ A/B report records the selected model and ~$0.006 spend. Headless
231/231, Python 94/94.
Add the alley walls, warm lantern, shutters, dumpster, AC clutter and door
step for the Joke Shop. Build the studio stone facade, bulb canopy, glossy
black car, stanchions and barriers. Rebuild only the world container when the
venue changes, so the guard and UI survive; add per-venue navigation checks.
Capture gallery shots 12/13, update the concept mapping and docs. Headless
231/231, Python 87/87, screenshot gate 13 good/0 bad. The vision-judge retry is
tracked separately because the API key is invalid.
Blender library
- Rebuild the face: sclera, iris and upper lid per eye, brow ridge, cheekbones,
nose bridge and tip, ears with lobes, split upper/lower lips.
- Give each hand four fingers plus a thumb and end every sleeve in a cuff ring.
- Raise CHAR_MAX_VERTS 4000 -> 5000 (16 avatars still cost ~6.9 ms of a 50 ms
frame), decimate the body to 26 %; characters are now 3,974-5,026 tris and
3,744-4,807 verts. Rebuild all 8 variants and re-bake their AO maps.
Capture tooling
- Add a `portrait` stage to capture_realism.gd: freeze one NPC, park the crowd
out of frame, stand the subject on open sidewalk, turn it into the camera and
put the guard on the sunlit side. New --portrait-type=N picks the role.
- Add the npc_portrait_day concept and map the shot to it, so avatar detail is
scored against a portrait instead of a wide street frame.
Flashlight
- Drop the fake additive beam (Materials.light_beam + the cone mesh): on GL
Compatibility a first-person camera inside the cone renders it as a filled
disc with a bright far ring. Keep the tight spot (14 deg, 20 m, energy 4.2)
and its headless test; the visible shaft is now planned behind a Forward+
preset.
Docs and task tracking
- Move the finished realism follow-ups (ground exposure, flashlight, NPC detail,
judge re-check) to DONE.md and expand TODO.md with step-by-step plans for NPC
gear fit, venue geometry plus the committed gallery, and the Forward+ preset.
- Record the decisions, gotchas and portrait recipe in MEMORY.md; update the
Blender contract, the concept README and the capture section of README.md.
- Refresh the gallery captures affected by the avatar rebuild.
Checks: headless 228/228, Python 86/86, blender_assets audit exit 0,
scene_concepts audit all GOOD, screen_designs compare 11 good / 0 bad,
smoke_npc_budget OK, boot clean (only the known headless mesh warning).
- add the v2 runtime picks: per-phase skyline panoramas, brick, sidewalk
pavers, dry/wet asphalt, awning canvas, and their normal maps
- add the realism tooling and evidence: tools/scene_concepts.py, capture_realism.gd,
concept frames, reports/sheets, and tests
- retune phase lighting/fog/exposure, add sun yaw, brighten the night glow,
add the flashlight beam mesh, and rebuild the planter/tree assets
- regenerate all eleven staged screenshots from the current build; compare is
11 good / 0 bad
- finish the visual-alignment closeout in DONE/TODO, update architecture,
blender-assets, PLAN, and ignore rules (editable Blender sources untracked)
- keep the judge retry and document the verified default judge model
Verification: headless 227/227; Python tooling 86/86; check_screenshots 11/11.
zwitschi