motion-ui
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by downloading and reading the source code of external project templates. \n
- Ingestion points: The agent downloads and reads source code from user-provided template links (SKILL.md). \n
- Boundary markers: The instructions do not define clear boundaries or provide warnings to disregard instructions found within the external project files. \n
- Capability inventory: The skill involves writing and modifying code for the Remotion framework and interacting with media generation tools. \n
- Sanitization: There is no mention of sanitizing or validating the content of the downloaded source code. \n- [EXTERNAL_DOWNLOADS]: The skill performs downloads of external project files based on user-supplied links. \n
- Evidence: "download and verify its exact original project, read its source" (SKILL.md).
Audit Metadata