# MCP The daemon exposes the knowledge base over MCP as a search tool (`ask_knowledge`) and as per-file resources. ## Tool: `ask_knowledge` Agents call `ask_knowledge` mid-task to retrieve context on demand: ```json { "tool": "ask_knowledge", "query": "PR review standards" } ``` Returns `` sections ordered by relevance score. The tool only registers when `qmd` is present and the index is ready — if it's missing from the tool list, check `bento daemon logs` for `knowledge: index ready`. ## Resources Each markdown file in `wiki/`, `blueprints/`, and `recipes/` is registered as an MCP resource at startup: ``` wiki://git-workflow blueprint://feature-evaluation-rubric recipe://add-agent ``` Any MCP client can browse and read these resources directly, without calling `ask_knowledge`. ## Automatic injection During the `retrieve` phase of a pipeline run with a retrieving [method](/knowledge-base/modes) (`bm25`/`vector`), the daemon queries the index and injects matching chunks into the agent prompt. Tune retrieval depth via [`knowledge.retrieval.topK`](/knowledge-base/config).