c4-level3-component

Pass

Audited by Gen Agent Trust Hub on May 21, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of architectural documentation and Mermaid diagram templates. It contains no executable scripts, network operations, or sensitive file access.
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to scan the local codebase for specific folder names (e.g., services/, controllers/) to synthesize diagrams.
  • Ingestion points: Scanning of local project directory (SKILL.md).
  • Boundary markers: Not explicitly defined in the prompt instructions.
  • Capability inventory: No active capabilities or subprocess calls are defined within the skill itself.
  • Sanitization: Not applicable as the skill performs structural analysis rather than content execution.
Audit Metadata
Risk Level
SAFE
Analyzed
May 21, 2026, 08:11 AM
Security Audit — agent-trust-hub — c4-level3-component