model-supply-chain
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The file is purely informational, providing an assessment methodology and grep commands for analyzing external code bases. It contains no operational scripts, dynamic execution blocks, or network exfiltration patterns.
- [PROMPT_INJECTION]: The static detector flag is a false positive. Phrases like "Ignore previous instructions" and "You are now..." appear strictly as illustrative examples within the documentation for Step 4d (MCP Indirect Prompt Injection via Tool Results), describing what to look for when auditing external tool responses, rather than attempting to manipulate the analyzing agent.
Audit Metadata