Script Compose

Takes user input (story idea, pitch, or formatted screenplay) and either captures it verbatim or rewrites it into proper screenplay format before storing it as a canvas note. Built specifically for the filmmaking canvas workflow. Won't trigger on file drops alone, you have to explicitly ask it to work. Once the script is captured, you can tell it to split the screenplay into shot notes capped at 15 seconds each and extract production anchors like characters, locations, and dialogue that downstream skills need for image and video generation. Preserves dialogue word-for-word and keeps shots as close to the 15 second model limit as the material allows. The split between this and the actual video generation skills is intentional, it hands off clean planning data instead of trying to do everything at once.

GitHub stars: 293
last updated: Jun 9, 2026

View on GitHub

Install

$ npx -y skills add utopai-research/pai-pro --skill script-compose --agent claude-code

Installs into .claude/skills of the current project.

Files

Run only on explicit user intent, never on file drop. Dropped text/PDF already exists as a note (data.body) and mirror (./assets/notes/<note_id>.md).

Defaults: a 30s beat is one moment; match the input language; with characters, prefer meaningful dialogue and let narration support rather than carry the scene, but let wordless action carry a beat instead of packing every one with dialogue.

Stop at script capture, shot notes, and anchor extraction. Route multi-stage work back to story-to-video-workflow.

Capture target duration when observable: