qinghu-douyin-social
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources (Douyin video comments and user profiles), which could theoretically contain malicious instructions designed to influence the agent's behavior.
- Ingestion points: Data is fetched from the Douyin platform via tools like
douyin_video_commentsanddouyin_user_profileinto the agent's context. - Boundary markers: The skill instructions do not define specific delimiters or "ignore instructions" prompts to isolate the ingested social media content.
- Capability inventory: The skill maintains the capability to perform network operations to the vendor's API endpoint (
iqinghu.com). - Sanitization: No explicit sanitization, filtering, or escaping of the natural language content retrieved from the social media platform is mentioned.
- [SAFE]: The skill adheres to best practices for secret management by instructing the agent to retrieve API tokens from environment variables (
QINGHU_TOKEN,QHKIT_TOKEN) or by requesting them directly from the user, avoiding hardcoded credentials. - [SAFE]: The skill includes a mandatory user authorization workflow for invoking tools and a standardized method for reporting consumption costs, ensuring transparency in API usage.
Audit Metadata