data-integration-assistant
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The documentation provides instructions to download official ThinkingData SDKs and tools from official sources (e.g., download.thinkingdata.cn, npm, and PyPI). These references are informative and point to legitimate product resources.
- [COMMAND_EXECUTION]: The skill includes diagnostic commands (e.g., curl health checks) and standard installation commands (e.g., npm install, pip install) as part of the integration documentation for users. These commands are benign and standard for the product's use case.
- [DATA_EXFILTRATION]: No unauthorized data exfiltration patterns were detected. The skill is designed to help users report analytics data to the ThinkingData platform as part of their application's legitimate analytics functionality.
- [PROMPT_INJECTION]: The instructions do not contain any patterns attempting to override agent safety guidelines or extract system prompts. The constraints provided are related to the accuracy of documentation-based answers.
Audit Metadata