generated: '2026-09-20'
method: searched
status: published
source: https://curatorsearch.com/mcp
description: >-
  Official hosted MCP server for CuratorSearch. Stateless Streamable HTTP (POST) endpoint on the
  provider's primary domain, no key required. Live tools/list retrieved anonymously and saved
  verbatim to mcp/curatorsearch-mcp-tools-list.json.
deployment:
  mode: remote
  endpoint: https://curatorsearch.com/mcp
  auth: none
  verified: probed
server:
  name: curatorsearch
  transport: http
  url: https://curatorsearch.com/mcp
  authentication: none
  notes: >-
    POST-only stateless Streamable HTTP. tools/list returned HTTP 200 with four read-only tools,
    each carrying a real inputSchema (and outputSchema). All tools are annotated readOnlyHint:true,
    destructiveHint:false, idempotentHint:true.
tools:
  - name: search_jobs
    description: >-
      Search live museum, gallery and curatorial job listings, gathered daily from institutions' own
      career pages. Returns up to 20 roles, newest first, each with its curatorsearch.com page.
    read_only: true
    input:
      - {name: query, type: string, required: false, description: "Free text matched against title, institution and location."}
      - {name: category, type: string, required: false, description: "One field of work (enum: Leadership, Curatorial, Education & Public Programmes, Archives, Collections, Conservation, Intern & Fellowship, Open Call & Residency, Production, Other)."}
      - {name: region, type: string, required: false, description: "One world region (enum: North America, UK & Ireland, Europe, Asia-Pacific, Middle East, Latin America, Africa, International)."}
      - {name: limit, type: integer, required: false, description: "How many roles to return, 1 to 20. Defaults to 10."}
    source_operation: openapi/curatorsearch-openapi.json#searchJobs
  - name: get_job
    description: >-
      One listing in full: description, stated salary, deadline, and the institution's original
      posting URL. Takes the slug from a curatorsearch.com/opportunities/<slug> URL.
    read_only: true
    input:
      - {name: slug, type: string, required: true, description: "The listing's slug, or the whole curatorsearch.com/opportunities/<slug> URL."}
    source_operation: openapi/curatorsearch-openapi.json#getJob
  - name: institution_jobs
    description: Every role one museum or gallery currently has open on the board, by institution name.
    read_only: true
    input:
      - {name: institution, type: string, required: true, description: "The institution's name, e.g. 'Philadelphia Museum of Art'. Near-misses resolve."}
    source_operation: null
    notes: No dedicated public REST operation; server-side filtered view of the live board.
  - name: salary_transparency
    description: >-
      How many live museum and curatorial postings state a salary, and the median advertised figure
      among those that do -- the CuratorSearch Curatorial Pay Archive headline number, computed live.
    read_only: true
    input: []
    source_operation: openapi/curatorsearch-openapi.json#salaryArchive
    notes: Computed aggregate over the salary archive; not a 1:1 projection of the raw REST array.
x-evidence:
  fetched: '2026-09-20'
  url: https://curatorsearch.com/mcp
  method: POST tools/list (jsonrpc 2.0)
  http_status: 200
  content_type: application/json