tiangong-kb-edu-search

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

• Third-party content exposure detected (medium risk: 0.65). The required workflow ingests request_file/input_file JSON via the wrapper’s --input "$REQUEST_FILE" path, and that JSON can contain arbitrary free-text fields (e.g., query/filter/fields) authored by an outsider, which the CLI then reads and forwards into the LLM context during retrieval.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's runtime default invokes "npx @tiangong-ai/cli@latest", which fetches and executes a remote npm package at runtime and is required for the CLI operations, so external code is downloaded and executed during runtime.