Review local video files by extracting representative frames, building contact sheets, inferring scene labels conservatively from visible content, and renaming clips with a safe CSV-backed workflow. Use when the user asks what is in a folder of videos, wants Codex to inspect video素材 from screenshots, wants clips renamed or organized by visual content, or needs a mapping table before batch renaming `.mov`, `.mp4`, `.m4v`, `.avi`, `.mkv`, or `.webm` files.
Use this skill to turn a folder of opaque video filenames into a reviewable, traceable rename plan. The workflow is deliberately conservative: inspect frames first, describe only what is visible, and keep every rename reversible through a CSV mapping.
scripts/build_review_assets.py.sheet_*.png files with view_image to inspect representative frames.scripts/apply_rename_csv.py --apply only after the labels look right.kitchen_person_cookingnight_city_lightsstreet_interview_two_people001_scene_subject_originalstem.ext.Run:
python3 scripts/build_review_assets.py "/path/to/videos"
Important outputs:
index.csv: one row per clip with duration, dimensions, and thumbnail pathsheet_01.png, sheet_02.png, ...: contact sheets for multimodal reviewthumbs/: extracted representative framesUse --output-dir when the source folder should stay clean or when the folder is large.
Create a CSV with at least these columns:
old_namenew_nameChinese headers 原文件名 and 新文件名 also work.
Recommended process:
视频重命名对照表.csv or video_rename_plan.csv.Preview first:
python3 scripts/apply_rename_csv.py "/path/to/视频重命名对照表.csv"
Apply after verification:
python3 scripts/apply_rename_csv.py "/path/to/视频重命名对照表.csv" --apply
The script validates collisions and uses a two-phase rename so swaps and overlapping targets remain safe.
~/.codex/skills, run the scripts from the repo root with relative paths as shown above.${CODEX_HOME:-$HOME/.codex}/skills/video-content-renamer.