Skip to content

Darwin MCP server

so.darwin/darwindarwin.sostreamable-httpv0.13.1Scanned

Browse the agentic web with Darwin: search, communicate, authenticate, and act through agents.

9 tools: 6 can delete or cancel things and 3 only read.

Tools: 6 delete, 0 irreversible, 0 write, 3 read.
  • Reads private data · 1
  • Reads text others write · 7
  • Can send data out · 1
  • 8 findings

Tools listed without sign-in. Remote posture

Anlyon alternative

6 of these tools change something. Run those as governed actions.

Declare what an action changes, and Anlyon puts it under a shared limit, binds every approval to the exact request, and grades the outcome. It works on any HTTPS API you can describe as an action.

Anlyon Radar does not run, proxy or govern this MCP server. A tool your own code calls directly is not routed through Anlyon and is not governed by it.

Import an OpenAPI spec →

What the Darwin MCP server can do

The Darwin MCP server is published by darwin.so, a domain the MCP Registry verified. It has 9 tools. It can delete or cancel things (6 tools: navigate, act, search, live-search and 2 more). 3 tools only read.

1 tool reads private data (get_thread); 7 tools bring in text other people can write (navigate, act, search and 4 more); 1 tool can send data out of your system (route). On its own it has all three parts of a prompt-injection exfiltration path: private data, text others write, and a way out.

The fixed checks matched 8 findings (credentials-mention and annotation-conflict). Each one comes with the text it matched.

Its remote endpoint lists its tools without sign-in.

Last scanned Nov 9, 2026, version 0.13.1.

Tools

Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.

9 tools

Deletes or cancels 6Claude Code: deny
  • actStart work from searchId or explicit targets, or continue a thread using threadId and a typed message.Reads text others write3 findings

    Why it is in this group

    • The description's first verb is "start": a word for a change.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.

    Findings

    • Info

      The input schema at anyOf[0].properties.searchToken.description mentions "credential".

      Matched text

      inputSchema.anyOf[0].properties.searchToken.description: "...at search and never substitute an OAuth credential."

      What this check looks for
    • Info

      The input schema at anyOf[1].properties.searchToken.description mentions "credential".

      Matched text

      inputSchema.anyOf[1].properties.searchToken.description: "...at search and never substitute an OAuth credential."

      What this check looks for
    • Info

      The input schema at anyOf[2].properties.searchToken.description mentions "credential".

      Matched text

      inputSchema.anyOf[2].properties.searchToken.description: "...at search and never substitute an OAuth credential."

      What this check looks for
    Description, as the server wrote it
    Start work from searchId or explicit targets, or continue a thread using threadId and a typed message. Authentication and payment responses use this tool. Preserve idempotencyKey on retries. Use get_act_request with the returned actRequestId for verified thread state; acceptance is not completion or settlement.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
  • executeSend an explicit execute turn.Reads text others writeNo findings

    Why it is in this group

    • The name "execute" says "execute": a word for a change.
    • The description's first verb is "send": a word for sending, and what is sent cannot be taken back.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Send an explicit execute turn. Omit threadId to start a thread; supply the returned threadId to continue. Use a stable clientMessageId for retries. This tool never automatically reclassifies the chosen mode. Execution and sensitive actions retain their authorization and approval requirements. Read progress with get_thread.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
  • navigateFind agents, check live task fit when supported, coordinate work or open direct agent conversations.Reads text others writeNo findings

    Why it is in this group

    • The name "navigate" says "navigate": a word for a change.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Find agents, check live task fit when supported, coordinate work or open direct agent conversations. Clarification and follow-up discussion stay in the current thread. Auto is the default. Send a message with a stable clientMessageId; continue using threadId and its issued continuation token when present. Explicit modes: search, live_search, execute, route. Live Search requires enabled provider support; unavailable features fail explicitly. Execution, budget changes and approvals require exact API-issued scope and authority; ordinary text never authorizes payment. Read progress with get_thread. Typed controls use conversationId and exact saved revisions or quotes: quote, authorize_budget, execute, stop, pause, release_budget, accept_step, refresh, advance_turn, resume, claim, revise, update, or message_step. Consequential controls require OAuth and their existing approval fields; do not infer consent from a turn.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
  • routeSend an explicit route turn.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The description's first verb is "send": a word for sending, and what is sent cannot be taken back.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe description's first verb is "send": it sends or publishes something to other people.
    Description, as the server wrote it
    Send an explicit route turn. Omit threadId to start a thread; supply the returned threadId to continue. Use a stable clientMessageId for retries. This tool never automatically reclassifies the chosen mode. Execution and sensitive actions retain their authorization and approval requirements. Read progress with get_thread.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
Only reads 3Show 3 read-only tools
  • get_accountRead the authenticated person’s safe Darwin account context and authorized AIs.1 finding

    Why it is in this group

    • The name "get_account" says "get": a word for reading.

    Findings

    • Info

      The description mentions "credentials".

      Matched text

      description: "...authorized AIs. Excludes email, phone, credentials, API keys, and application administrati..."

      What this check looks for
    Description, as the server wrote it
    Read the authenticated person’s safe Darwin account context and authorized AIs. Excludes email, phone, credentials, API keys, and application administration.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • get_threadRead an owned Navigate or agent thread using threadId.Reads private dataReads text others write1 finding

    Why it is in this group

    • The name "get_thread" says "get": a word for reading.

    What it can reach

    • private-dataThe name "get_thread" says "get" and "thread": it reads messages.
    • private-dataThe description's first sentence says "read" and "thread": it reads messages.
    • untrusted-inputThe name "get_thread" says "get" and "thread": it reads messages, which other people can write.
    • untrusted-inputThe description's first sentence says "read" and "thread": it reads messages, which other people can write.

    Findings

    • Info

      The description mentions "credentials".

      Matched text

      description: "...or a guest Navigate thread. Agent guest credentials remain in host custody. Use the returne..."

      What this check looks for
    Description, as the server wrote it
    Read an owned Navigate or agent thread using threadId. Pass the issued continuation token for a guest Navigate thread. Agent guest credentials remain in host custody. Use the returned pagination cursor for earlier messages. Reading never starts work or grants approval. Legacy thread arguments remain accepted.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • show_uiDisplay a read-only view of exact agent matches, a current approval, authentication or payment request, or a completed output.No findings

    Why it is in this group

    • The name "show_ui" says "show": a word for reading.
    Description, as the server wrote it
    Display a read-only view of exact agent matches, a current approval, authentication or payment request, or a completed output. Use only presentation tokens and request IDs returned by Darwin. Displaying a view never approves work, authenticates an account, pays, or changes state. Hosts without UI receive the same readable data.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false

This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.

Exfil surface

Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.

7 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 1 outbound tool breaks every one.

Reads private data1
Reads files, mail, messages, databases, secrets, calendars or repositories.
Reads text others write7
Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
Can send data out1
Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
deny to break every path
{
  "permissions": {
    "deny": [
      "mcp__darwin__route"
    ]
  }
}

Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.

Remote posture

What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.

Tools listed without sign-in

Looked at Nov 20, 2026, 19:21 UTC, unchanged since Nov 20, 2026. No credentials were sent and no tool was called. What is looked at.

Reachable
yes
HTTP status of an anonymous initialize
200
tools/list served without credentials
yes
Tools that change things, listed without credentials
6
Protocol version
2025-06-18
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
https://mcp.darwin.so/.well-known/oauth-protected-resource/mcp
Authorization servers
https://darwin.so/api/customer/auth
Legacy HTTP+SSE transport
no
Dynamic client registration
yes
Client ID metadata documents
no
  • HighThe whole server

    The endpoint served tools/list without credentials, and the list has 6 tools that change things (4 destructive, 2 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://mcp.darwin.so/mcp: tools/list returned 9 tools to an anonymous client, including navigate (destructive), run (destructive), search (destructive), live-search (destructive), build (write) and 1 more

    What this check looks for

Posture changes

  1. Nov 20, 2026
    • Reachable: no to yes
    • HTTP status of an anonymous initialize: none to 200
    • tools/list served without credentials: no to yes
    • Tools that change things, listed without credentials: 0 to 6
    • Protocol version: none to 2025-06-18
    • RFC 9728 protected resource metadata: none to https://mcp.darwin.so/.well-known/oauth-protected-resource/mcp
    • Authorization servers: none to https://darwin.so/api/customer/auth
    • Dynamic client registration: not observed to yes
    • Client ID metadata documents: not observed to no
  2. Nov 13, 2026
    • Reachable: yes to no
    • HTTP status of an anonymous initialize: 200 to none
    • tools/list served without credentials: yes to no
    • Tools that change things, listed without credentials: 6 to 0
    • Protocol version: 2025-06-18 to none
    • RFC 9728 protected resource metadata: https://mcp.darwin.so/.well-known/oauth-protected-resource/mcp to none
    • Authorization servers: https://darwin.so/api/customer/auth to none
    • Dynamic client registration: yes to not observed
    • Client ID metadata documents: no to not observed
  3. Nov 6, 2026
    • Reachable: no to yes
    • HTTP status of an anonymous initialize: none to 200
    • tools/list served without credentials: no to yes
    • Tools that change things, listed without credentials: 0 to 6
    • Protocol version: none to 2025-06-18
    • RFC 9728 protected resource metadata: none to https://mcp.darwin.so/.well-known/oauth-protected-resource/mcp
    • Authorization servers: none to https://darwin.so/api/customer/auth
    • Dynamic client registration: not observed to yes
    • Client ID metadata documents: not observed to no
  4. Oct 29, 2026
    • Reachable: yes to no
    • HTTP status of an anonymous initialize: 401 to none
    • WWW-Authenticate on 401: yes to not observed
    • resource_metadata in WWW-Authenticate: https://mcp.darwin.so/.well-known/oauth-protected-resource/mcp to none
    • RFC 9728 protected resource metadata: https://mcp.darwin.so/.well-known/oauth-protected-resource/mcp to none
    • Authorization servers: https://darwin.so/api/customer/auth to none
    • Dynamic client registration: yes to not observed
    • Client ID metadata documents: no to not observed

Findings

Matches of fixed checks, with the text they matched.

2 warn6 info
  • Warnsearch

    The server marks "search" destructive (destructiveHint: true), and its name says "search", a word for reading.

    Matched text

    annotations: {"readOnlyHint":false,"destructiveHint":true,"idempotentHint":false,"openWorldHint":true}; The name "search" says "search": a word for reading.

    What this check looks for
  • The server marks "live-search" destructive (destructiveHint: true), and its name says "search", a word for reading.

    Matched text

    annotations: {"readOnlyHint":false,"destructiveHint":true,"idempotentHint":false,"openWorldHint":true}; The name "live-search" says "search": a word for reading.

    What this check looks for
  • Infoact

    The input schema at anyOf[0].properties.searchToken.description mentions "credential".

    Matched text

    inputSchema.anyOf[0].properties.searchToken.description: "...at search and never substitute an OAuth credential."

    What this check looks for
  • Infoact

    The input schema at anyOf[1].properties.searchToken.description mentions "credential".

    Matched text

    inputSchema.anyOf[1].properties.searchToken.description: "...at search and never substitute an OAuth credential."

    What this check looks for
  • Infoact

    The input schema at anyOf[2].properties.searchToken.description mentions "credential".

    Matched text

    inputSchema.anyOf[2].properties.searchToken.description: "...at search and never substitute an OAuth credential."

    What this check looks for
  • Infosearch

    The input schema at anyOf[1].properties.searchToken.description mentions "credential".

    Matched text

    inputSchema.anyOf[1].properties.searchToken.description: "...at search and never substitute an OAuth credential."

    What this check looks for
Show 2 more findings
  • The description mentions "credentials".

    Matched text

    description: "...or a guest Navigate thread. Agent guest credentials remain in host custody. Use the returne..."

    What this check looks for
  • The description mentions "credentials".

    Matched text

    description: "...authorized AIs. Excludes email, phone, credentials, API keys, and application administrati..."

    What this check looks for

History

Changes to its tool definitions and to its remote endpoint, newest first.

  1. PostureRemote posture changed. Reachable: no -> yes; HTTP status of an anonymous initialize: none -> 200; tools/list served without credentials: no -> yes (and 6 more).View diff
    • Reachable: no to yes
    • HTTP status of an anonymous initialize: none to 200
    • tools/list served without credentials: no to yes
    • Tools that change things, listed without credentials: 0 to 6
    • Protocol version: none to 2025-06-18
    • RFC 9728 protected resource metadata: none to https://mcp.darwin.so/.well-known/oauth-protected-resource/mcp
    • Authorization servers: none to https://darwin.so/api/customer/auth
    • Dynamic client registration: not observed to yes
    • Client ID metadata documents: not observed to no
  2. PostureRemote posture changed. Reachable: yes -> no; HTTP status of an anonymous initialize: 200 -> none; tools/list served without credentials: yes -> no (and 6 more).View diff
    • Reachable: yes to no
    • HTTP status of an anonymous initialize: 200 to none
    • tools/list served without credentials: yes to no
    • Tools that change things, listed without credentials: 6 to 0
    • Protocol version: 2025-06-18 to none
    • RFC 9728 protected resource metadata: https://mcp.darwin.so/.well-known/oauth-protected-resource/mcp to none
    • Authorization servers: https://darwin.so/api/customer/auth to none
    • Dynamic client registration: yes to not observed
    • Client ID metadata documents: no to not observed
  3. PostureRemote posture changed. Reachable: no -> yes; HTTP status of an anonymous initialize: none -> 200; tools/list served without credentials: no -> yes (and 6 more).View diff
    • Reachable: no to yes
    • HTTP status of an anonymous initialize: none to 200
    • tools/list served without credentials: no to yes
    • Tools that change things, listed without credentials: 0 to 6
    • Protocol version: none to 2025-06-18
    • RFC 9728 protected resource metadata: none to https://mcp.darwin.so/.well-known/oauth-protected-resource/mcp
    • Authorization servers: none to https://darwin.so/api/customer/auth
    • Dynamic client registration: not observed to yes
    • Client ID metadata documents: not observed to no
  4. PostureRemote posture changed. Reachable: yes -> no; HTTP status of an anonymous initialize: 401 -> none; WWW-Authenticate on 401: yes -> not observed (and 5 more).View diff
    • Reachable: yes to no
    • HTTP status of an anonymous initialize: 401 to none
    • WWW-Authenticate on 401: yes to not observed
    • resource_metadata in WWW-Authenticate: https://mcp.darwin.so/.well-known/oauth-protected-resource/mcp to none
    • RFC 9728 protected resource metadata: https://mcp.darwin.so/.well-known/oauth-protected-resource/mcp to none
    • Authorization servers: https://darwin.so/api/customer/auth to none
    • Dynamic client registration: yes to not observed
    • Client ID metadata documents: no to not observed

Install

Pinned to version 0.13.1. Values in angle brackets are yours to fill in.

claude mcp add --transport http darwin https://mcp.darwin.so/mcp
claude mcp add --transport http darwin https://mcp.darwin.so/mcp

Run in a terminal.

Claude Code permissions

Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.

.claude/settings.json
{
  "permissions": {
    "allow": [
      "mcp__darwin__get_thread",
      "mcp__darwin__show_ui",
      "mcp__darwin__get_account"
    ],
    "ask": [],
    "deny": [
      "mcp__darwin__navigate",
      "mcp__darwin__act",
      "mcp__darwin__search",
      "mcp__darwin__live-search",
      "mcp__darwin__route",
      "mcp__darwin__execute"
    ]
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as darwin. Review it before you use it.

Capability tags behind 7 rules
  • mcp__darwin__navigate untrusted-input
  • mcp__darwin__act untrusted-input
  • mcp__darwin__search untrusted-input
  • mcp__darwin__live-search untrusted-input
  • mcp__darwin__route untrusted-input, outbound
  • mcp__darwin__execute untrusted-input
  • mcp__darwin__get_thread private-data, untrusted-input

Pin

Keep the decision: fail CI when the tools change, follow every change, or show a badge.

Pin in CI

Fail a build when the tools you reviewed are not the tools the server serves today.

curl -s "https://radar.anlyon.com/api/v1/servers/so.darwin/darwin/verify?hash=c1f3d69f914d49c2e965101ef70c31df1cb4cbfc96fcfb250118e4c6e4cc4739" | jq -e '.changed == false'

Pinned hash c1f3d69f914d, from Nov 9, 2026, 07:18 UTC.

Watch

Get told when its tools change: a new tool, a new description, a new class.

https://radar.anlyon.com/servers/so.darwin/darwin/changes.xml

An Atom feed. Works in any feed reader or Slack's RSS app.

Badge

Anlyon Radar badge for so.darwin/darwin
[![Anlyon Radar](https://radar.anlyon.com/badge/so.darwin/darwin.svg)](https://radar.anlyon.com/servers/so.darwin/darwin)

Source: MCP Registry · repository. Synced Nov 10, 2026, 01:18 UTC.

Darwin MCP server: questions and answers

Answered from the scan above, not from the publisher's description.

Is the Darwin MCP server safe?

No scan can promise that. What its definitions show: 9 tools, 6 of which can change, send or delete something, and 8 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.

Can the Darwin MCP server delete data?

Yes. 6 tools can delete, cancel or revoke something: navigate, act, search, live-search, route and execute.

Can the Darwin MCP server do things that cannot be undone?

No tool is classed as irreversible: none of them sends, pays, publishes or merges, going by their definitions.

What tools does the Darwin MCP server have?

9 tools: 6 destructive, 3 read. navigate, act, search, live-search, route, execute, get_thread, show_ui and 1 more.

Does the Darwin MCP server need sign-in or OAuth?

Its remote endpoint listed its tools to an anonymous client, without sign-in.

Does the Darwin MCP server send data to outside services?

1 tool can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): route.

Has the Darwin MCP server changed its tools?

Not since Anlyon Radar started tracking it. Last scanned Nov 9, 2026.