cross-check-legal-advice
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is restricted to instructional markdown content and prompt templates for legal research. It does not perform any automated actions, file operations, or network requests.
- [NO_CODE]: There are no scripts (Python, JavaScript, Shell, etc.) or binary files included in the skill. This eliminates primary attack vectors such as remote code execution, persistence, or privilege escalation.
- [INDIRECT_PROMPT_INJECTION]: The skill templates define ingestion points for untrusted external data, specifically contract clauses pasted into the agent context (SKILL.md). While this creates an injection surface, the skill has a zero-capability inventory (no subprocess calls, file-write, or network operations), ensuring that even if an injection were attempted through a processed document, there are no tools available for the agent to execute malicious actions. The templates do not define explicit boundary markers or sanitization logic, but the risk remains negligible due to the absence of executable capabilities.
Audit Metadata