pr-review-expert

Pass

Audited by Gen Agent Trust Hub on Apr 2, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) due to its processing of untrusted external data.\n
  • Ingestion points: Reads pull request metadata and code diffs directly from external sources using gh pr view and gh pr diff.\n
  • Boundary markers: The instructions do not define delimiters or protective instructions to isolate untrusted PR text from the agent's logic.\n
  • Capability inventory: The skill has the capability to execute shell commands (e.g., npm test, pytest, gh) and perform network requests via curl.\n
  • Sanitization: There is no evidence of sanitization or validation of the PR content before it is processed for review.\n- [COMMAND_EXECUTION]: Uses established command-line utilities including gh, glab, grep, jq, and wc to interact with repositories and analyze code. These operations are standard and necessary for a pull request review tool.\n- [DATA_EXFILTRATION]: Performs network requests to official Jira and Linear API endpoints to verify issue status. These communications use environment variables for authentication and target well-known, trusted services.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 2, 2026, 02:36 PM