merge
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill uses output from git log and git diff to generate titles and descriptions for new pull requests and issues.
- Ingestion points: git commit messages and diff output (SKILL.md).
- Boundary markers: Absent.
- Capability inventory: GitHub CLI (gh) execution for creating issues and PRs (SKILL.md).
- Sanitization: Absent.
- [COMMAND_EXECUTION]: The skill explicitly directs the agent to unset authentication-related environment variables (GITHUB_TOKEN and GH_TOKEN) when calling the GitHub CLI to ensure correct credential usage in the execution environment.
Audit Metadata