| name | mofa-frame |
| version | 0.1.0 |
| description | Extract a video frame at a specific timestamp. Triggers: extract frame, 提取画面, screenshot video, capture frame, 截图, grab frame, video frame at, 视频截图. |
| requires_bins | ffmpeg |
| requires_env | null |
mofa-frame
Extracts a single frame from a video at a specific timestamp. Returns the frame as a PNG image.
Usage
Use after mofa_youtube — read the transcript, identify interesting moments, then extract frames at those timestamps.
mofa frame --video-path ./video.mp4 --timestamp 120 --label "key diagram"
Workflow
mofa_youtube downloads video + transcribes → returns transcript.md + video path
- Agent reads transcript, identifies key moments with timestamps
- Agent calls
mofa_frame for each interesting timestamp
- Each frame is delivered to chat
Output
Single PNG file, resized to 800px wide. Filename includes timestamp and optional label.