draft-release-notes
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from git commit logs and GitHub API responses which may contain adversarial instructions embedded in PR titles or descriptions.
- Ingestion points: Shell commands
git logandgh api repos/:owner/:repo/releases/generate-notesinSKILL.mdprovide external content to the agent. - Boundary markers: The skill provides a specific markdown template and narrative style guidelines to constrain the output format.
- Capability inventory: The skill has the capability to read git history and write modifications to the local
CHANGELOG.mdfile. - Sanitization: No explicit sanitization or escaping of the fetched commit/PR text is performed prior to interpolation.
Audit Metadata