architecture-diagram

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via the codebase metadata it processes.
  • Ingestion points: The skill ingests untrusted data from the codebase structure via the cartograph:codebase-explorer agent in Step 1.
  • Boundary markers: There are no boundary markers or instructions telling the agent to ignore potentially malicious content within the extracted module or package names.
  • Capability inventory: The skill uses the mcp__claude_ai_Mermaid_Chart__validate_and_render_mermaid_diagram tool to render the output, which involves processing the potentially poisoned Mermaid syntax.
  • Sanitization: The instructions do not specify sanitization or escaping for the names extracted from the codebase before they are interpolated into the generated Mermaid flowchart syntax.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 10:05 AM
Security Audit — agent-trust-hub — architecture-diagram