generated: '2026-08-04'
method: derived
source:
- mcp/lifemine-mcp.yml
- openapi/lifemine-content-openapi.yml
- openapi/lifemine-careers-openapi.yml
- graphql/lifemine-graphql.yml
description: >
  Binds every candidate MCP tool to the OpenAPI operation that backs it, so each tool inherits a
  REAL input schema instead of a guessed one. Confidence is `high` throughout because both the
  tools and the operations were derived from the same fetched documents in this repo — there is no
  provider-published tool list to reconcile against. The GraphQL column is unverifiable: the
  endpoint is live but public introspection is disabled, so no field-level mapping is asserted.
surfaces:
  openapi:
  - file: openapi/lifemine-content-openapi.yml
    base_url: https://lifeminetx.com/wp-json
    operations: 268
    gated: false
    note: Read operations anonymous; write operations return 401 rest_forbidden without a WordPress credential.
  - file: openapi/lifemine-careers-openapi.yml
    base_url: https://boards-api.greenhouse.io/v1/boards/lifeminetx
    operations: 12
    gated: false
  graphql:
    endpoint: https://lifeminetx.com/graphql
    gated: true
    reason: >
      WPGraphQL public introspection disabled. Root type confirmed as RootQuery, but no schema
      could be read, so no GraphQL field is mapped in any row below.
  mcp:
    url: null
    published: false
    gated: null
    note: No MCP server exists. Tools are candidates derived from the OpenAPIs.
crosswalk:
- tool: list_news
  category: content
  rest: [getWpV2Posts]
  binding: rest
  confidence: high
- tool: get_news_item
  category: content
  rest: [getWpV2PostsById]
  binding: rest
  confidence: high
- tool: list_news_categories
  category: content
  rest: [getWpV2Categories]
  binding: rest
  confidence: high
- tool: search_site
  category: content
  rest: [getWpV2Search]
  binding: rest
  confidence: high
  note: >
    /wp/v2/search returns typed cross-resource stubs (id, title, url, type, subtype), not full
    objects — a follow-up get_news_item / get_page call is needed for the body.
- tool: list_pages
  category: content
  rest: [getWpV2Pages]
  binding: rest
  confidence: high
- tool: get_page
  category: content
  rest: [getWpV2PagesById]
  binding: rest
  confidence: high
- tool: list_people
  category: people
  rest: [getWpV2Team]
  binding: rest
  confidence: high
- tool: get_person
  category: people
  rest: [getWpV2TeamById]
  binding: rest
  confidence: high
- tool: list_people_groups
  category: people
  rest: [getWpV2TeamCategories]
  binding: rest
  confidence: high
- tool: list_media
  category: content
  rest: [getWpV2Media]
  binding: rest
  confidence: high
- tool: list_jobs
  category: careers
  rest: [listJobs]
  binding: rest
  confidence: high
- tool: get_job
  category: careers
  rest: [getJob]
  binding: rest
  confidence: high
- tool: list_departments
  category: careers
  rest: [listDepartments]
  binding: rest
  confidence: high
- tool: list_offices
  category: careers
  rest: [listOffices]
  binding: rest
  confidence: high
mcp_only: []
mcp_only_note: >
  Empty by construction — the candidate tool list was derived FROM the REST surface, so no tool
  can exist without a backing operation. A published LifeMine MCP server would be the thing that
  could introduce composites; none exists.
rest_only:
- capability: WordPress content authoring
  operations_approx: 139
  examples: [createWpV2Posts, updateWpV2PostsById, deleteWpV2PostsById, createWpV2Media, patchWpV2PagesById]
  reason: >
    Write operations across posts, pages, media, blocks, templates, menus and taxonomies. Excluded
    from the tool list because they require a WordPress credential that is not issued to the
    public.
- capability: Block editor and site-building internals
  examples: [getWpV2BlockTypes, getWpV2Templates, getWpV2TemplateParts, getWpV2GlobalStylesById, getWpV2Widgets, getWpV2Navigation]
  reason: Editor/theme plumbing with no agent-facing value against a marketing site.
- capability: Site health diagnostics
  examples: [getWpSiteHealthV1TestsHttpsStatus, getWpSiteHealthV1TestsLoopbackRequests, getWpSiteHealthV1DirectorySizes]
  reason: Administrative diagnostics; auth-gated in practice.
- capability: WordPress Abilities API
  examples: [getWpAbilitiesV1Abilities, getWpAbilitiesV1Categories]
  reason: >
    Registered on the host and genuinely agent-oriented — a capability registry an agent could
    discover and run — but returns 401 rest_forbidden anonymously, so nothing can be bound to it.
    This is the single most interesting unrealised agent surface on this provider.
- capability: Users, comments, statuses, oembed
  examples: [getWpV2Users, getWpV2Comments, getWpV2Statuses, getOembed10Embed]
  reason: Readable but editorially unused on this site; omitted to keep the tool set meaningful.
- capability: Greenhouse education reference data
  operations: [listEducationDegrees, listEducationDisciplines, listEducationSchools, getBoard, listSections, getSection, getDepartment, getOffice]
  reason: >
    Application-form picklists and singletons that are Greenhouse platform data rather than
    LifeMine information.
- capability: Third-party plugin admin namespaces
  namespaces: [yoast/v1, wordfence/v1, wpe/cache-plugin/v1, wpe_sign_on_plugin/v1]
  reason: >
    Real registered routes on the live host, but vendor control planes rather than LifeMine's
    content surface. Excluded from openapi/lifemine-content-openapi.yml entirely; the complete
    verbatim route index is preserved at openapi/_original/lifemine-wp-json-index.json.
coverage:
  tools_named: 14
  tools_bound: 14
  mcp_only: 0
  rest_ops_total: 280
  rest_ops_in_specs:
    content: 268
    careers: 12
  rest_ops_with_a_tool: 14
  rest_read_ops_with_a_tool_pct: ~11
  note: >
    Low coverage is the correct outcome, not a gap: 139 of the 268 content operations are
    credentialed writes, and most of the remaining reads are block-editor and theme plumbing. The
    14 tools cover essentially all of the publicly meaningful read surface — news, pages, people,
    media, search, jobs, departments and offices.