address-github-comments

Pass

Audited by Gen Agent Trust Hub on Feb 21, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [Indirect Prompt Injection] (LOW): The skill possesses an attack surface for indirect prompt injection as it ingests untrusted data from GitHub PR comments.
  • Ingestion points: External comments are fetched via gh pr view --comments in sub-skills/1-inspect-comments.md.
  • Boundary markers: None. The skill does not instruct the agent to treat comment content as data rather than instructions.
  • Capability inventory: The skill allows for local file modifications (applying fixes) and external communication (gh pr comment).
  • Sanitization: There is no mention of sanitizing or escaping the content of the comments before processing.
  • [Command Execution] (SAFE): The skill uses the standard GitHub CLI (gh) for its operations. While it executes shell commands, they are strictly related to the declared purpose of the skill (authenticating, viewing PRs, and commenting).
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 21, 2026, 10:28 AM