semi-intel-platform
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions define a legitimate process for patent intelligence gathering and reporting. It correctly identifies the author as Patsnap and specifies the use of official Patsnap MCP data sources. The guidelines prohibit data fabrication and include necessary legal disclaimers, adhering to standard security and ethics best practices for AI agents. No malicious behavior such as prompt injection, obfuscation, or unauthorized access was detected.\n- [INDIRECT_PROMPT_INJECTION]: The skill possesses a standard data-processing surface, as its primary purpose is to ingest user-provided data to generate HTML reports. This is an inherent part of its functionality and not a malicious finding.\n
- Ingestion points: User-provided objects, conditions, and files are processed as specified in the SKILL.md sections describing analysis tasks.\n
- Boundary markers: No specific delimiters are mandated for external data, but the skill emphasizes strict adherence to Patsnap MCP data.\n
- Capability inventory: The skill generates single-file HTML dashboards, charts, and brief reports based on the analysis results.\n
- Sanitization: There are no explicit instructions for sanitizing user-provided inputs before they are rendered in the generated HTML output.
Audit Metadata