remotion-best-practices
Pass
Audited by Gen Agent Trust Hub on Jul 21, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill documents patterns that create an attack surface for indirect prompt injection through external data ingestion.
- Ingestion points: The skill provides examples of fetching data from remote URLs to define video properties, specifically in
rules/calculate-metadata.md(viaprops.dataUrl),rules/compositions.md(viaprops.videoId),rules/import-srt-captions.md(external .srt files), andrules/lottie.md(external JSON assets). - Boundary markers: The provided code examples lack boundary markers or explicit instructions to the agent to ignore potential commands embedded within the fetched data.
- Capability inventory: The ingested data is used to dynamically calculate composition metadata (duration, width, height) and set component props, which influences the final render output and agent state.
- Sanitization: There is no evidence of data sanitization or schema validation in the examples before the fetched data is merged into the application state.
Audit Metadata