code-reviewer

Pass

Audited by Gen Agent Trust Hub on Oct 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The scripts/review_helper.py script executes external command-line tools such as bandit, radon, pylint, and safety to perform code analysis.
  • The script uses subprocess.run with list-based arguments, which is a secure implementation that avoids shell interpolation risks.
  • This functionality is essential to the skill's purpose as an automated review helper.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted user-provided source code, creating a surface for indirect prompt injection attacks.
  • Ingestion points: The agent reads user-provided code files and references the included security pattern documentation (examples/security_patterns.md).
  • Boundary markers: The instructions do not specify strict delimiters (e.g., XML tags) for isolating user code from the agent's internal instruction context.
  • Capability inventory: The skill can execute shell commands via local scripts and generate code suggestions, which could be manipulated by malicious input in the code being reviewed.
  • Sanitization: The skill provides documentation to help the agent identify insecure patterns, but the agent itself relies on its core safety training to avoid obeying instructions embedded within the code it is auditing.
  • [DYNAMIC_EXECUTION]: The skill performs runtime execution of external analysis tools based on user-supplied file paths. This is a standard pattern for developer tooling and does not involve the execution of arbitrary remote code or unverified binaries.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 16, 2026, 04:23 PM