gen-release-notes

Pass

Audited by Gen Agent Trust Hub on May 9, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill relies on shell com m and s like git log, git diff, and gh pr list to extract repository metadata. It uses $AR G UMENTS to specify the starting tag for the comparison.
  • [IN D I RECT_PROMPT_IN J ECTION]: The skill proces s es untrusted com m it mes s ages and pul l request titles to generate des c ription s for the changelog. This pres ent s a potential surface for in d i rec t prom p t inj ec t ion where malicious con t en t in a com m it coul d at t em p t to influence the agen t's out p ut for m a t t ing or categorization logic.
  • In g es t ion poin t s: Data is sourced from local git logs (git log) and ext er n al pul l request sum m a ries (gh pr list).
  • Boun d a ry mar k er s: The in s t r uc t ion s do not def ine spec i f ic boun d a ries or ig n or e-in s t r uc t ion war n ing s for the proces s ed com m it data.
  • Capabil i t y inv en t or y: The skill has ac c es s to shell exec u t ion (git, gh) and local file sy s t em mod i f ic a t ion (CHANGELOG.md).
  • San i t iz a t ion: Com m it mes s ages and PR data ar e proces s ed wit h out exp l ic it san i t iz a t ion or fil t er ing of in s t r uc t ion s.
Audit Metadata
Risk Level
SAFE
Analyzed
May 9, 2026, 12:42 PM