chinese-legal-commercial
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external legal documents (commercial contracts, NDAs, SaaS agreements) which may contain adversarial instructions.
- Ingestion points: The skill routes natural language requests and processes contract content for review, risk reporting, and summarization.
- Boundary markers: There are no explicit delimiters or instructions provided to distinguish between user requests and the content of the legal documents being analyzed.
- Capability inventory: The skill utilizes tools for local file system access, web verification, and document rendering.
- Sanitization: No sanitization or validation mechanisms are defined for the ingested legal text.
- [DATA_EXFILTRATION]: The skill contains instructions to read configuration data from local hidden directories.
- Evidence: It specifically directs the agent to read from
~/.claude/plugins/config/to extract existing user practice profiles and setup configuration.
Audit Metadata