code-review

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes untrusted data from pull request diffs, metadata, and repository guidance files while maintaining write capabilities.
  • Ingestion points: Untrusted data enters the agent context through gh pr view, gh pr diff, and repository-specific files such as AGENTS.md, CLAUDE.md, and contribution documentation.
  • Boundary markers: The instructions do not specify any delimiters or warnings to ignore embedded instructions within the reviewed code changes.
  • Capability inventory: The skill uses gh pr comment to post feedback, which could be exploited to perform actions on behalf of the agent if it obeys instructions embedded in the diff.
  • Sanitization: No sanitization, escaping, or validation of the ingested content is described in the skill logic.
  • [COMMAND_EXECUTION]: The skill executes local CLI tools to perform its core functions.
  • Evidence: The skill uses git (for local diffs and history) and gh (for GitHub pull request metadata and commenting).
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 10:19 PM
Security Audit — agent-trust-hub — code-review