patent-pre-evaluation-report
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied technical disclosures and external patent data to generate pre-evaluation reports. This ingestion surface is a standard requirement for the skill's primary function of patent analysis. 1. Ingestion points: Technical plan extraction (Step 1) and external patent data retrieval via novelty_fetch_patent_data (Step 4) in SKILL.md. 2. Boundary markers: The skill uses structural extraction logic; explicit boundary delimiters for untrusted data are not specified in the instructions. 3. Capability inventory: The skill utilizes the files.begin_write and files.append tools to output HTML files to the @session/reports/ directory. 4. Sanitization: The skill does not explicitly mention filtering of patent abstracts before inclusion in reports.
- [COMMAND_EXECUTION]: The skill invokes a sequence of specialized MCP tools (e.g., novelty_summary, novelty_search_agent) from the hub-mcp-gateway-novelty-search server to perform technical evaluation tasks. These tools are part of the vendor's official service offering.
- [EXTERNAL_DOWNLOADS]: The skill fetches patent and literature data from official Patsnap domains, including open.zhihuiya.com and open.patsnap.com, which is necessary for real-time search and retrieval capabilities.
Audit Metadata