system-understanding-agent

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill is a documentation tool that performs read-only analysis of a repository's structure and existing documentation.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external metadata files to generate its reports.
  • Ingestion points: .understand-anything/knowledge-graph.json and .understand-anything/diff-overlay.json (SKILL.md).
  • Boundary markers: The instructions explicitly command the agent not to hallucinate and to only use confirmed information (SKILL.md).
  • Capability inventory: The skill is restricted to writing markdown files to a specific local directory (.agile-v/phases/00-understanding/). It cannot execute shell commands or make network requests.
  • Sanitization: No explicit sanitization of the input JSON content is required by the instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 06:07 AM
Security Audit — agent-trust-hub — system-understanding-agent