metaclaw-evolving-agent
Warn
Audited by Snyk on Sep 12, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.30). The workflow intercepts and processes live user-agent chat conversations and downstream LLM interactions through a local proxy, which ingests arbitrary user-authored free text from client requests.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata