analyze-code

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted data (local source code) without sufficient boundary enforcement.
  • Ingestion points: The skill explicitly gathers existing code files from the current working directory to be analyzed by the 'consultant' agent.
  • Boundary markers: The instructions lack explicit delimiters (e.g., XML tags or unique separators) to help the LLM distinguish between the source code content and the analysis instructions.
  • Capability inventory: The sub-agent invokes a 'consultant CLI' and generates a markdown report for the user.
  • Sanitization: There is no evidence of content sanitization or instruction-aware filtering applied to the code files before they are provided to the LLM.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 01:33 AM