reviewing-pull-requests

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses local Python and Bash scripts (scripts/pr-manager.py and scripts/quality-gates.sh) to interact with the GitHub CLI (gh) and git. These scripts utilize argument lists in subprocess calls, which is a secure practice that prevents shell injection vulnerabilities when processing untrusted pull request data such as branch names, titles, and descriptions.\n- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection (Category 8) because its primary function is to ingest and analyze untrusted data.\n
  • Ingestion points: Untrusted data enters the agent context through PR titles, descriptions, and code diffs fetched via gh pr view and gh pr diff in the management scripts.\n
  • Boundary markers: The skill instructions do not specify any delimiters or safety prompts to ensure the agent ignores instructions that may be embedded within the PR content being reviewed.\n
  • Capability inventory: The skill possesses the ability to create/edit pull requests, assign reviewers, and update project boards via the gh tool.\n
  • Sanitization: While the scripts prevent shell-level injection, there is no natural language sanitization to filter or neutralize malicious instructions contained within the pull request data itself.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 09:58 AM