kami e8941d8ceb Stop labelling a model guess as a read tail ...

_annotate_speaker_methods stamped `tail`, the highest-trust provenance, on any
line whose speaker matched a present local_id, at gemma's confidence of 1.0. No
balloon was read. Three of three sampled two-character panels had the speakers
swapped, so a multi-character guess is now dropped to unknown, and a solo-panel
guess is kept as model_solo at 0.7.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

2026-08-11 23:23:56 +04:00

attic

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

caveats

Stop labelling a model guess as a read tail

2026-08-11 23:23:56 +04:00

decisions

Stop labelling a model guess as a read tail

2026-08-11 23:23:56 +04:00

scripts

Restore runtime deps: models, dots.tts, torchvision shadow

2026-08-11 03:28:26 +04:00

systemd

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

.gitignore

Restore runtime deps: models, dots.tts, torchvision shadow

2026-08-11 03:28:26 +04:00

AGENTS.md

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

AUDIT.md

Audit Phase 1: correctness and scheduling safety

2026-08-11 10:16:24 +04:00

bubble_detect.py

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

CLAUDE.md

Audit Phase 1: correctness and scheduling safety

2026-08-11 10:16:24 +04:00

collage.py

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

face_detect.py

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

HANDOFF.md

Record the script verifier failure on two-word names

2026-08-11 23:02:06 +04:00

identity-rework-task.md

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

JOURNAL.md

Stop labelling a model guess as a read tail

2026-08-11 23:23:56 +04:00

manga-recap-pipeline-spec.md

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

NEXT.md

Stop labelling a model guess as a read tail

2026-08-11 23:23:56 +04:00

plan.md

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

requirements.txt

Restore runtime deps: models, dots.tts, torchvision shadow

2026-08-11 03:28:26 +04:00

ROADMAP.md

Audit Phase 1: correctness and scheduling safety

2026-08-11 10:16:24 +04:00

session_manager.py

Audit Phase 1: correctness and scheduling safety

2026-08-11 10:16:24 +04:00

spec-correctness.md

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

spec-v3.md

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

start_workers.sh

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

test_vision_parse.py

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

transport.py

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

worker_crop.py

Split artifacts across per-class buckets, record the baseline run

2026-08-11 23:00:13 +04:00

worker_identity.py

Reconstruct repo from Claude Code + codex transcripts

2026-08-11 02:42:41 +04:00

worker_layers.py

Split artifacts across per-class buckets, record the baseline run

2026-08-11 23:00:13 +04:00

worker_render.py

Split artifacts across per-class buckets, record the baseline run

2026-08-11 23:00:13 +04:00

worker_scene.py

Audit Phase 1: correctness and scheduling safety

2026-08-11 10:16:24 +04:00

worker_script.py

Audit Phase 1: correctness and scheduling safety

2026-08-11 10:16:24 +04:00

worker_tts.py

Split artifacts across per-class buckets, record the baseline run

2026-08-11 23:00:13 +04:00

worker_vision.py

Stop labelling a model guess as a read tail

2026-08-11 23:23:56 +04:00

S

Description

No description provided

686 KiB

Languages

Python 97.1%

Shell 2.9%