volcengine-sdk-generator

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow ingests outsider-authored free text via the user’s query: scripts/rg_rank.py sends --query to the API Explorer search/all endpoint and then parses/returns highlighted title/abstract/content strings as candidate metadata that influence subsequent API selection/code generation.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The scripts (scripts/make_code.py, scripts/rg_rank.py, and scripts/sdk_info.py) call the remote API base https://api.volcengine.com/api/common at runtime (e.g., /explorer/api-swagger and /explorer/make-code) to fetch swagger and generated code which are injected into the agent output, so remote content directly controls the skill's produced prompts/code.