generated: '2026-09-01'
method: probed
source: >-
  Live unauthenticated JSON-RPC POST {"jsonrpc":"2.0","id":1,"method":"tools/list"} to each of the
  three endpoints below on 2026-09-01; every one answered HTTP 200 with a full tools[] array
  including inputSchema. Endpoints and registry names read from the provider's own
  https://www.pathwren.workers.dev/.well-known/mcp.json.
deployment:
  mode: remote
  endpoint: https://www.pathwren.workers.dev/mcp
  install: null
  package: null
  auth: none
  verified: probed
  note: >-
    Remote only. No npx/stdio package is published in any registry (see packages/), so the
    only way to reach this provider through MCP is the hosted Streamable-HTTP endpoint —
    which an agent can call right now with no install step and no key.
transport: streamable-http
protocol_versions:
- '2026-07-28'
- '2025-11-25'
- '2025-06-18'
- '2025-03-26'
- '2024-11-05'
registry:
  url: https://registry.modelcontextprotocol.io
  note: >-
    /.well-known/mcp.json mirrors the official MCP Registry record for each server verbatim
    under x-registry-record, schema and all.
servers:
- name: dev.workers.pathwren.www/ai-crawler-index
  title: AI Crawler Index
  version: 1.1.0
  endpoint: https://www.pathwren.workers.dev/mcp
  documentation: https://www.pathwren.workers.dev/mcp.html
  probe:
    method: tools/list
    http_status: 200
    tool_count: 6
    file: mcp/pathwren-ai-crawler-index-tools.json
  tools:
  - name: classify_user_agent
    required: [user_agent]
    description: Identify the crawler behind a raw User-Agent header — operator, category, robots.txt stance, verification method, cost of blocking.
  - name: lookup_crawler
    required: [slug]
    description: The full index record for one crawler; an unknown slug returns near matches.
  - name: list_crawlers
    required: []
    parameters: [category, operator, respects_robots_txt, verification, q, limit]
    description: The index, filtered or whole; the reply enumerates every valid filter value.
  - name: generate_robots_txt
    required: []
    parameters: [stance]
    description: A ready-to-paste robots.txt block for one of eight stances, with reasoning and the crawlers it names.
  - name: is_verified_crawler_ip
    required: [ip]
    description: Test an IPv4/IPv6 address against every operator-published crawler prefix; a miss is reported as a miss, not a forgery.
  - name: whats_changed
    required: []
    parameters: [since]
    description: Build time, per-source fetch state, which sources changed or are failing, and the dated changelog.
  annotations:
    note: >-
      Every tool carries readOnlyHint:true, destructiveHint:false, idempotentHint:true,
      openWorldHint:false in its own annotations block — the server declares itself
      side-effect free.
- name: dev.workers.pathwren.www/crawler-log-triage
  title: Crawler Log Triage
  version: 1.0.0
  endpoint: https://www.pathwren.workers.dev/mcp/triage
  documentation: https://www.pathwren.workers.dev/mcp-triage.html
  probe:
    method: tools/list
    http_status: 200
    tool_count: 5
    file: mcp/pathwren-crawler-log-triage-tools.json
  tools:
  - name: triage_log
    required: [log]
    parameters: [log, detail, limit]
  - name: find_impersonators
    required: [log]
  - name: summarize_by_operator
    required: [log]
  - name: robots_from_log
    required: [log]
    parameters: [log, stance]
  - name: waf_ruleset_from_log
    required: [log]
    parameters: [log, target, scope, action]
  note: No tool in common with the AI Crawler Index server — different input (a whole access log), different job.
- name: dev.workers.pathwren.www/agent-discovery-doctor
  title: Agent Discovery Doctor
  version: 1.0.0
  endpoint: https://www.pathwren.workers.dev/mcp/doctor
  documentation: https://www.pathwren.workers.dev/mcp-doctor.html
  probe:
    method: tools/list
    http_status: 200
    tool_count: 5
    file: mcp/pathwren-agent-discovery-doctor-tools.json
  tools:
  - name: check_discovery_documents
    required: [host]
    parameters: [host, groups, mcp_path]
  - name: explain_document
    required: []
    parameters: [name]
  - name: validate_llms_txt
    required: [text]
  - name: llms_txt_from_sitemap
    required: [xml]
    parameters: [xml, title, summary, max_per_section]
  - name: validate_agent_card
    required: [json]
resources:
- https://www.pathwren.workers.dev/data/agents.json
- https://www.pathwren.workers.dev/data/ua-regex.json
- https://www.pathwren.workers.dev/ip-ranges/all.json
- https://www.pathwren.workers.dev/status.json
authorization:
  required: false
  oauth_protected_resource_metadata: null
  oauth_authorization_server_metadata: null
  note: >-
    Probed anonymously: /.well-known/oauth-protected-resource, /.well-known/oauth-authorization-server
    and /.well-known/openid-configuration all return 404 with a JSON body. The provider documents
    this as intentional in /.well-known/mcp.json x-authorization — the server never issues a 401
    and never inspects an Authorization header.