Stats
Actions
Tags
Concat a cover still + main video + optional outro still into one 1080×1920 mp4, normalizing fps / SAR / audio sample rate / channel layout per leg so the join is seamless. Use whenever the user wants to "拼接封面 / 加片头 / 加片尾 / bookend / stitch cover and outro / 视频前后加封面图". Output: `<stem>_published.mp4` next to the main video.
How this skill is triggered — by the user, by Claude, or both
Slash command
/podcast-video-toolkit:stitch-bookendThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Wrap a clip with a cover still at the head and (optionally) an outro still at the tail. All legs are normalized in a single `filter_complex` so VFR / SAR / audio mismatches don't break the concat.
Wrap a clip with a cover still at the head and (optionally) an outro still at the tail. All legs are normalized in a single filter_complex so VFR / SAR / audio mismatches don't break the concat.
bash "${SKILL_DIR}/scripts/stitch.sh" <main.mp4> \
--cover <cover.jpg> \
[--outro <outro.png>] \
[--cover-secs 2] \
[--outro-secs 3.5] \
[-o <out.mp4>]
Default output: <main-dir>/<main-stem>_published.mp4.
<main.mp4> — reads fps, has-audio, duration. If no audio track, synthesizes anullsrc for the main leg.1080×1920, SAR=1, fps=30, yuv420p with letterbox/pillarbox padding on white.44100Hz stereo aac (silent for the still legs).+faststart, AAC 128k.cover_secs + main_dur + outro_secs (±0.1s), 1080×1920.If the post-stitch check fails the script exits non-zero.
npx claudepluginhub pierrelzw/podcast-video-toolkit --plugin podcast-video-toolkitCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.