address-code-review

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses git pull and gh api to fetch feedback and synchronize the repository state. These commands are standard for development workflows and used for their intended purpose.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it processes untrusted data from external sources.
  • Ingestion points: Reads review comments from review.jsonl at the repository root, GitHub Pull Request threads (inline and general) via the gh CLI, and local markdown documents.
  • Boundary markers: The instructions do not define clear delimiters or warnings to ignore embedded instructions within the review data.
  • Capability inventory: The skill has the ability to write to the file system (apply code changes), delete files (review.jsonl), execute shell commands (git pull, git commit), and perform network operations via the GitHub API (replying to and resolving comments).
  • Sanitization: There is no evidence of sanitization or validation of the input data before it is presented to the agent or used to generate code changes.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 07:55 AM