openseo-release-notes

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it fetches untrusted text from git commit logs and PR descriptions (Step 2) to draft release notes and guide subagent reviews (Step 4). * Ingestion points: PR bodies and titles fetched via gh pr view. * Boundary markers: None; the instructions do not specify any delimiters or safety warnings for the agent when processing external text. * Capability inventory: File system writes (package.json, release notes), shell command execution (git push, gh pr create, gh release create), and subagent spawning with repository access. * Sanitization: None; external text is interpolated directly into the release notes and review tasks.
  • [COMMAND_EXECUTION]: The skill executes shell commands using git, gh, and pnpm to perform repository management and automate the GitHub release lifecycle (Steps 1, 2, 5).
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 05:56 PM
Security Audit — agent-trust-hub — openseo-release-notes