patent
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted patent-related data from external JSON files, providing an attack surface where malicious instructions could be embedded within invention descriptions or patent claims.
- Ingestion points: Data enter the agent context through
invention.json,patents.json, andpatentability.jsonwhich are read by the analysis scripts in thescripts/directory. - Boundary markers: There are no instructions or delimiters defined to separate the processed data from the agent's execution instructions.
- Capability inventory: The skill scripts (e.g.,
patentability_scorer.py) read local files and produce markdown output that is returned to the agent's context, potentially influencing its subsequent actions. - Sanitization: The skill does not validate, sanitize, or escape the input content before rendering it into the final markdown reports for the agent.
Audit Metadata