muapi-director
Turn ONE topic, talking-head video, or photo into a finished Vox-style paper-collage explainer / ad video on the MuAPI platform (api.muapi.ai) + local ffmpeg — script, collage keyframes, motion, voice-over, music, captions, all automated. Three input modalities: a topic (B-roll), a talking-head video (A-roll mode), or a single photo of a person/product anchored into the collage (C-roll mode).
The English summary is being prepared. Showing the original description.
See how to installFiles included (25)
- SKILL.md2.7 KB
- .gitignore349 B
- AGENTS.md1.1 KB
- assets/demo.mp46.3 MB
- assets/youtube-thumbnail.jpg100.1 KB
- examples/sample-beats.json2.4 KB
- package.json818 B
- README.md13.1 KB
- README.zh.md10.8 KB
- references/beat-layer.md3.3 KB
- references/prompt-guide.md3.0 KB
- references/voices.md1.4 KB
- scripts/aroll_assemble.py2.6 KB
- scripts/aroll_clips.py4.5 KB
- scripts/asr_beats.py4.9 KB
- scripts/assemble.py6.4 KB
- scripts/audio.py3.7 KB
- scripts/clips.py5.8 KB
- scripts/croll_keyframes.py5.1 KB
- scripts/keyframes.py3.1 KB
- scripts/muapi_client.py9.0 KB
- scripts/provider.py5.5 KB
- scripts/style_bakeoff.py2.4 KB
- scripts/styles.py11.5 KB
- scripts/text_overlay.py4.6 KB
SKILL.md (original)
Review the instructions the agent will receive before you install.
MuAPI Director
Turn a topic, talking-head video, or photo into a finished Vox-style paper-collage video: a bold, punchy, narrated explainer/ad where each beat is a torn-paper collage poster that comes alive, with voice-over, music and captions. Powered by MuAPI (MUAPI_API_KEY) + local ffmpeg.
Input Modalities
- B-roll (topic → film): Pass a topic (e.g. "A brief history of Silicon Valley"). It drafts a script, builds keyframe posters, animates them, synthesizes narration/BGM, and assembles the video.
- A-roll (talking-head video → collage): Pass a presenter talking-head recording. It runs ASR, segments the transcript, and re-styles the video into paper-collage graphics while preserving facial features, lip-sync, and gestures frame-for-frame.
- C-roll (one photo → collage): Pass a selfie or product shot. The subject is cut out as an unedited photographic sticker, anchored into generated collage posters per beat, and animated.
Prerequisites
MUAPI_API_KEYset in your environment (get one from https://muapi.ai).ffmpeg+ffprobeinstalled on system path.- Python 3 with
Pillow(pip install pillow).
Workflow Steps
B-roll Mode (Topic)
- Write
out/<project>/beats.jsoncontaining topic, language, and beat map. python3 scripts/style_bakeoff.py out/<project>(Pick theme from candidates).python3 scripts/keyframes.py out/<project>(Generates collage posters).python3 scripts/clips.py out/<project>(Animates posters into motion clips).python3 scripts/audio.py out/<project>(Generates TTS narration and instrumental BGM).python3 scripts/assemble.py out/<project>(Muxes audio, ducking, captions, watermark ->final.mp4).
A-roll Mode (Talking Head)
python3 scripts/asr_beats.py out/<project> <source.mp4>python3 scripts/aroll_clips.py out/<project>python3 scripts/aroll_assemble.py out/<project>
C-roll Mode (Photo Anchor)
- Add
"mode": "croll","anchor_photo": "photo.png","croll_subject": "portrait"|"product"tobeats.json. python3 scripts/croll_keyframes.py out/<project>python3 scripts/clips.py out/<project>python3 scripts/audio.py out/<project>python3 scripts/assemble.py out/<project>
Reviews
No reviews yet. Tell others what you thought.
Comments
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.