The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx to download and run the yuque-mcp package from the npm registry. As Yuque is a well-known documentation service, this is a standard operational requirement but involves executing remote code.
[COMMAND_EXECUTION]: The installation and setup instructions require the agent or user to execute several shell commands, including npx yuque-mcp setup and configuration commands for various MCP clients.
[INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by ingesting document content and search results from the Yuque API into the agent's context.
Ingestion points: The agent retrieves untrusted text data through tools such as yuque_get_doc, yuque_search, and yuque_get_note as specified in the skill documentation.
Boundary markers: Absent. The skill does not provide the agent with specific delimiters or instructions to treat the retrieved content as data distinct from its own instructions.
Capability inventory: The agent possesses significant write capabilities, including yuque_create_doc, yuque_update_doc, yuque_create_book, and yuque_update_toc, which could be exploited if the agent follows instructions hidden within retrieved documents.
Sanitization: Absent. There is no mention of sanitizing or filtering the content fetched from the Yuque knowledge base before it is processed by the AI.

yuque-mcp-server