gs-review-pr-reply

Pass

Audited by Gen Agent Trust Hub on Nov 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands using the GitHub CLI (gh) and git to retrieve pull request data and post reviews. It adheres to security best practices by utilizing the --input flag with a temporary file for the gh api POST request in Step 5. This method prevents shell syntax injection vulnerabilities that can occur when passing complex JSON payloads as string arguments in the command line.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data which presents a surface for indirect prompt injection.
  • Ingestion points: External data is ingested from pull request diffs (gh pr diff), existing GitHub comments (gh api pulls/.../comments), and previous /gs-review reports within the conversation history.
  • Boundary markers: The skill does not explicitly define boundary markers or use "ignore embedded instructions" delimiters when processing this data.
  • Capability inventory: The skill possesses network capabilities (GitHub API interactions) and file system access (writing payload files outside the repository).
  • Sanitization: The skill includes instructions to the agent to manually review and remove secrets, tokens, or private paths from the draft. The primary mitigation is a mandatory human-in-the-loop checkpoint where the user must approve the final draft before it is posted to GitHub.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 20, 2026, 07:02 PM