chinese-legal-procurement
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXPOSURE]: The skill accesses user configuration profiles stored in
~/.claude/plugins/config/and~/.codex/legal-zh/procurement-legal/. This access is intended to retrieve and store user-specific legal practice settings and setup profiles. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze procurement documents and legal queries, which serves as a potential attack surface for instructions hidden within external data.
- Ingestion points: Natural language legal requests and procurement-related documents (e.g., bidding files, contracts) provided by the user in SKILL.md.
- Boundary markers: The instructions do not define specific delimiters or "ignore embedded instructions" warnings for the content being analyzed.
- Capability inventory: The skill uses tools for local file access to load workflow instructions and potentially utilizes web verification and document rendering tools.
- Sanitization: There are no explicit instructions for the agent to sanitize or escape instructions that might be embedded in user-provided legal documents.
Audit Metadata