requesting-code-review

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill uses a template for the code-reviewer subagent that ingests untrusted data from the repository (e.g., implementation descriptions and requirements) through placeholders. While this constitutes a potential indirect prompt injection surface, it is necessary for the skill's primary function of code review and is mitigated by the use of read-only analysis tools.
  • Ingestion points: Placeholders {WHAT_WAS_IMPLEMENTED}, {PLAN_OR_REQUIREMENTS}, and {DESCRIPTION} in code-reviewer.md are used to inject external context.
  • Boundary markers: There are no explicit delimiters used to wrap the placeholder content in the prompt.
  • Capability inventory: The subagent's tools (git diff, get_symbols, deep_dive, fast_refs) are confined to repository observation and code structure analysis.
  • Sanitization: The instructions do not define specific sanitization or escaping logic for the placeholder data.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 03:01 PM
Security Audit — agent-trust-hub — requesting-code-review