yuque-mcp-server

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

• Third-party content exposure detected (high risk: 0.75). The required workflow uses Yuque MCP tools like yuque_search / yuque_get_doc to fetch Yuque document content at runtime, and that document body/snippets are outsider-authored free text (other users’ collaborative docs) that can be injected into the agent’s LLM context via tool results.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill launches the MCP server via npx (e.g., "npx yuque-mcp") which fetches and executes remote npm package code at runtime (repository referenced: https://github.com/yuque/yuque-mcp-server), so external code is executed and relied upon to operate the skill.