requesting-code-review

Pass

Audited by Gen Agent Trust Hub on Apr 24, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute standard version control commands including git rev-parse, git log, and git diff to identify the range of changes for review. These operations are appropriate for a developer-oriented skill focused on code analysis.
  • [PROMPT_INJECTION]: The skill uses a template in code-reviewer.md that incorporates user-provided or document-sourced text through placeholders like {DESCRIPTION} and {PLAN_OR_REQUIREMENTS}. This creates a surface for indirect prompt injection where instructions hidden in the code description or planning documents could attempt to bias the reviewer subagent's output.
  • Ingestion points: Variables substituted into the code-reviewer.md template from external plan documents and user input.
  • Boundary markers: None identified; input text is interpolated directly into the instructions.
  • Capability inventory: The skill allows the execution of git commands and provides an assessment that could influence the agent's workflow (e.g., deciding whether to proceed with development).
  • Sanitization: No validation or sanitization is performed on the content of the placeholders before they are passed to the reviewer agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 24, 2026, 10:13 AM
Security Audit — agent-trust-hub — requesting-code-review