legal-global-router
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates as a specialized router for legal tasks, directing user queries to appropriate sub-skills or platform commands based on topic and jurisdiction.
- [SAFE]: Robust safety guidelines are included, explicitly prohibiting the agent from providing direct legal advice, signing contracts, or bypassing attorney review. The 'NEVER DO THESE' and 'UNIVERSAL RULES' sections provide clear operational boundaries.
- [SAFE]: The skill utilizes local jurisdictional reference files (e.g., uk-law.md, eu-law.md) to provide relevant legal context without external network calls.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it is designed to ingest and analyze untrusted external data (contracts, NDAs, and playbooks). However, this risk is mitigated by the skill's primary function as an analysis tool for human review and the lack of dangerous automated capabilities (such as file-writing or network exfiltration). The mandatory output header requiring attorney review serves as a human-in-the-loop safety mechanism.
Audit Metadata