dev
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill establishes an attack surface by instructing the agent to ingest untrusted data from the repository's git history (
git log) and existing markdown files to generate documentation. - Ingestion points: Project file contents (files.md, task.md) and git commit history.
- Boundary markers: None specified.
- Capability inventory: File system write access for documentation updates and git command execution.
- Sanitization: None specified for commit message processing.
- [SAFE]: The skill references numerous official resources and setup guides from well-known services including Convex (docs.convex.dev), React (react.dev), Resend (resend.com), and Vercel (vercel.com). These references are documentation-focused and align with standard developer best practices.
Audit Metadata