nestjs-git-commit-pr-message
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements best practices for Git automation, requiring explicit user authorization for remote actions and providing clear decision boundaries between different task-specific skills.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from Git logs and remote pull request/issue descriptions. 1. Ingestion points: Remote prose and PR/issue bodies via Git/GitHub inspection (SKILL.md). 2. Boundary markers: Explicit instructions to treat remote prose as 'untrusted evidence' (SKILL.md). 3. Capability inventory: Execution of Git commands (git commit, git push) and GitHub CLI operations (gh pr create) for repository mutation. 4. Sanitization: Clear policy to 'never' treat remote content as instructions and to extract facts only.
- [SAFE]: The workflow includes a mandatory check for sensitive content in staged changes (e.g., API keys, private keys, authenticated URLs) using patterns defined in references/sensitive-content-and-remote-safety.md. It instructs the agent to stop the process and report the file category without echoing the secret values, preventing accidental exposure in logs or commits.
Audit Metadata