performance-testing-review-ai-review

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves as an instructional guide and implementation playbook for setting up code review workflows. It utilizes legitimate security and quality tools such as SonarQube, CodeQL, and Semgrep.
  • [COMMAND_EXECUTION]: The Python and shell examples demonstrate the use of subprocesses and CLI commands to orchestrate analysis tools. These are standard practices for the described DevOps automation tasks.
  • [DATA_EXFILTRATION]: Integration examples follow security best practices by using environment variables (e.g., GITHUB_TOKEN, OPENAI_API_KEY) and GitHub Secrets for authentication, rather than hardcoding credentials.
  • [PROMPT_INJECTION]: The skill exhibits an Indirect Prompt Injection attack surface, which is a common design consideration for code review agents.
  • Ingestion points: The system processes untrusted code diffs and PR descriptions (via $ARGUMENTS and script variables).
  • Boundary markers: The examples use markdown headers (e.g., Modified Code:) but do not include explicit instructions for the AI to ignore embedded commands within the diffs.
  • Capability inventory: The skill's demonstrated capabilities include running static analysis tools and posting comments back to pull requests.
  • Sanitization: The provided code templates do not demonstrate input sanitization or escaping for the interpolated content. Users implementing this blueprint should include robust delimiters and sanitization logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 09:30 PM
Security Audit — agent-trust-hub — performance-testing-review-ai-review