generated: '2026-08-01' method: derived source: mcp/mindmaze-mcp.yml + openapi/mindmaze-content-openapi.yml status: candidate important: >- MindMaze Therapeutics operates NO MCP server and NO GraphQL surface. This crosswalk therefore binds the CANDIDATE tool proposal in mcp/mindmaze-mcp.yml to the real OpenAPI operations that would back it, so each proposed tool inherits a real inputSchema instead of a guessed one. It is not a record of a published agent surface, and no MCPServer pointer is wired in apis.yml. surfaces: openapi: file: openapi/mindmaze-content-openapi.yml operations: 17 gated: false note: All 17 operations verified anonymous HTTP 200 on 2026-08-01. graphql: endpoint: null note: /graphql returns 404 on every MindMaze host. No GraphQL surface exists. mcp: url: null note: No hosted or remote MCP server. /.well-known/mcp.json returns 404 on every host. crosswalk: - tool: list_mindmaze_news category: disclosure rest: [listPosts] binding: rest confidence: high note: >- Inherits the full parameter set of GET /wp/v2/posts - page, per_page, search, after, before, order, orderby, slug, _fields, categories, tags. - tool: get_mindmaze_news_item category: disclosure rest: [getPost] binding: rest confidence: high - tool: list_mindmaze_pages category: content rest: [listPages] binding: rest confidence: high - tool: get_mindmaze_page category: content rest: [getPage] binding: rest confidence: high - tool: search_mindmaze_content category: search rest: [searchContent] binding: rest confidence: high - tool: list_mindmaze_categories category: taxonomy rest: [listCategories] binding: rest confidence: high note: >- A tool implementation should call this first to resolve disclosure category ids rather than hardcoding them. - tool: list_mindmaze_media category: content rest: [listMedia] binding: rest confidence: high note: mime_type=application/pdf isolates investor presentations and financial reports. mcp_only: [] mcp_only_note: >- Empty by construction - every candidate tool was derived from an OpenAPI operation, so no tool can exist without a backing REST operation. rest_only: - operations: [getApiIndex, getNamespaceIndex, listTaxonomies, listContentTypes, listStatuses] capability: discovery reason: >- Self-describing metadata. Useful to a client at build time, not to an agent at call time. - operations: [getMediaItem] capability: content reason: Single-record media fetch; the list operation already returns source_url. - operations: [getCategory, listTags, getTag] capability: taxonomy reason: >- Redundant with list_mindmaze_categories for agent purposes; only one tag term is registered. - operations: [getOEmbed] capability: embedding reason: Presentation-layer helper, not an information-retrieval capability. coverage: tools_named: 7 tools_bound: 7 mcp_only: 0 rest_operations_total: 17 rest_operations_with_a_tool: 7