The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or security risks were identified. The skill follows standard practices for integrating Stitch and Remotion tools.
[EXTERNAL_DOWNLOADS]: The skill fetches screenshot assets from Stitch URLs and utilizes the Remotion skills library. These are well-known and trusted sources.
Evidence: screenshot.downloadUrl, htmlCode.downloadUrl from the Stitch API.
Evidence: Reference to github.com/remotion-dev/skills and remotion.dev for transitions and components.
[COMMAND_EXECUTION]: The skill instructs the agent to use standard package management and build tools for project setup and rendering.
Evidence: Use of npm create video@latest, npm install @remotion/transitions, and npx remotion render for building the video composition.
[PROMPT_INJECTION]: The skill processes user-provided screen data, such as titles and descriptions, to generate video overlays and narration scripts.
Ingestion points: [stitch_prefix]:get_screen provides titles and descriptions from the project.
Boundary markers: Not explicitly specified in the prompts.
Capability inventory: Shell access (Bash), file writing, and Node.js execution (npm/npx).
Sanitization: Not specified, but the processing is limited to generating UI text and scripts within the Remotion framework.

stitch-remotion