linkfox-zhihuiya-description-data
Pass
Audited by Gen Agent Trust Hub on Aug 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests patent description and specification data from the external Zhihuiya database through the
zhihuiya/descriptionDataAPI. This external content is untrusted and could contain malicious instructions. - Ingestion points: External patent text is fetched by
scripts/zhihuiya_description_data.pyand presented to the agent. - Boundary markers: The skill instructions lack explicit boundary markers or "ignore instructions" delimiters to prevent the agent from accidentally executing instructions contained within patent descriptions.
- Capability inventory: The skill possesses capabilities for network access (via scripts), filesystem writes (caching and logging in the
/linkfox/directory), and shell command execution. - Sanitization: There is no evidence of sanitization or structural filtering applied to the retrieved technical descriptions to mitigate injection risks.
- [COMMAND_EXECUTION]: The skill relies on the execution of Python scripts (
zhihuiya_description_data.pyandonboarding.py) via the shell to process API requests and manage user environment settings.
Audit Metadata