pr
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: The skill uses standard development tools and workflows for managing branches and pull requests without any detected malicious intent.
- [EXTERNAL_DOWNLOADS]: The skill utilizes
bunx oxfmtto format code, which involves fetching a package from a public registry. This is consistent with standard development practices for the Remotion project. - [INDIRECT_PROMPT_INJECTION]: The skill has a potential attack surface when polling GitHub PR comments for Vercel preview links to include in the PR body. 1. Ingestion points: PR comments fetched via
gh pr viewor similar polling (SKILL.md). 2. Boundary markers: No specific delimiters are mentioned for parsing comment text. 3. Capability inventory:git push,gh pr create,gh pr edit,bun run(SKILL.md). 4. Sanitization: The skill significantly reduces risk by requiring the agent to write PR content to temporary files and use the--body-fileargument, which prevents shell injection from untrusted input.
Audit Metadata