MsNBBc3R 3dbee8f4aa

fix(skill-creator): preserve recorded token counts in benchmarks (#7033) ...

* fix(skill-creator): preserve recorded token counts in benchmarks

Read timing.json token measurements independently of grading duration and retain recorded zero values. Keep the existing character-count fallback for missing measurements and add benchmark/report regression coverage.

* fix(skillscan): distinguish token metrics from credentials

Recognize typed token counts, numeric benchmark summaries, and signed deltas so skill-creator documentation passes deterministic review. Preserve credential names, string values, duplicate-key detection, and cloud-token findings; add regression coverage and document the boundary.

---------

Co-authored-by: Willem Jiang <willem.jiang@gmail.com>

2026-10-18 20:44:39 +08:00

..

skill_loader.py

test: fix six Windows-host failures across sandbox, read_file, lark shim, skill review, and extension checks (#6787)

2026-10-14 06:19:43 +08:00

test_github_deep_research.py

fix(github-research): encode repository file paths in content requests (#7018)

2026-10-18 07:31:03 +08:00

test_image_generation.py

feat(skills): support OpenAI-compatible image generation (#5389)

2026-09-13 18:07:58 +08:00

test_music_generation.py

feat: MiniMax provider for image/video/podcast skills + new music-generation skill (#3437)

2026-06-08 22:04:38 +08:00

test_podcast_generation.py

fix(podcast): make Volcengine voices configurable (#5156)

2026-09-03 22:10:55 +08:00

test_skill_creator_benchmark.py

fix(skill-creator): preserve recorded token counts in benchmarks (#7033)

2026-10-18 20:44:39 +08:00

test_video_generation.py

fix(video): forward --aspect-ratio into the Gemini Veo request (#5388)

2026-09-13 18:09:15 +08:00