- Home
- Skills
- Image & Video
- AI Music Video Clip Pipeline
AI Music Video Clip Pipeline
Plan, index and assemble a song-synced AI clip library (lip-sync vs beat-sync), with a validator and budget rules.
$5
Works with the AI Agents you already use
About this skill
A finished song, no footage. The video will be assembled in an editor from many short generated clips, and the song is the only audio track: generated clips are muted (or generated silent). Everything below exists to make that edit fast, consistent, and affordable.
- The master audio is the clock. All timing is in seconds from the start of the final mixed song file. Never time against a demo or a re-export.
- Clips are silent. Any generated audio is discarded; lip-sync clips are driven by the isolated vocal stem or line audio but exported muted.
- Calibrate before you produce. No strong-generator spend until the calibration stills are approved (see section 5).
- Every clip lands in the index before it is used in an edit. No index row, no clip.
- Budget is a ceiling, not a target. Stop and ask when a phase hits its cap.
- Consent and rights. Only original characters, or real people who have given written consent. No copyrighted characters, logos, or lyrics you do not own.
Tight lip-sync clip (lip) | Loose beat-sync clip (beat) | |
|---|---|---|
| Purpose | Character visibly sings one line | Movement, mood, cutaways, B-roll |
| Length | The line + 0.3-0.5 s handles each side (typically 2-6 s) | 4-15 s, generated long, trimmed in edit |
| Timing | Bound to one lyric line (lyric_id) | Bound to a section or bar range, any slice usable |
| Driver | Line audio / vocal stem, or performance reference | Tempo (BPM) and energy prompt only |
| Framing | Face readable: CU, MCU, MS | Anything |
| Failure cost | High (visible mouth errors) | Low (trim around bad frames) |
Rules of thumb:
- Lip-sync only the lines the viewer must see sung: hooks, first line of each verse, the emotional peak. Typically 20-35% of lines; the rest is beat clips.
- Generate one lip clip per line, not per verse. Short clips sync better and are cheap to redo.
- Beat clips should have no mouth movement or the mouth turned away/obscured, so they never contradict the vocal.
- Ask beat clips for a clear motion accent every bar (step, turn, light flash) so the editor can snap cuts to the beat grid.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Video Skills.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 19 days ago
- Software
Python