gs-find-bad-comments

Pass

Audited by Gen Agent Trust Hub on Nov 20, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data in the form of code comments and git diff content, which could potentially contain malicious instructions designed to influence the agent's reporting behavior.
  • Ingestion points: The skill reads file content and git diffs directly from the local disk as specified in SKILL.md (Step 1).
  • Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard instructions embedded within the comments it is auditing.
  • Capability inventory: The skill is restricted to reading files and executing specific git metadata commands. It does not have the capability to write to the filesystem or perform network requests.
  • Sanitization: There is no evidence of sanitization or filtering of the comment text before it is processed or included in the final report.
  • [COMMAND_EXECUTION]: The skill utilizes several standard git commands to perform its auditing tasks.
  • Evidence: It uses git symbolic-ref and git remote show origin to determine the repository's default branch, and git log -L and git blame to trace the history of specific code segments. These operations are benign and necessary for the skill's intended functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 20, 2026, 09:35 AM