MCP Server for Z.AI - A Model Context Protocol server that provides AI capabilities
Darwin is installed from its publisher's own source and answers where it runs, so this marketplace is not in the path of a single call. There is no address here to send one to, and a panel that pretended otherwise would be showing you an answer we made up. Install it and call it from your own client — the Installation tab has the entry for each one.
What it does
Search public AI capabilities and coordinate approved work through Darwin's Search to Act interface.
Quickstart
# 1 — add it to Claude Code
claude mcp add --scope user --transport http darwin-mcp https://mcp.darwin.so/mcp
# 2 — ask your agent for something these tools do
# search, show_darwin_view, open_darwin and 13 more
Darwin costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.
Collected from a public index. Nobody has claimed this account, so nothing here was written by its author — claim it if it is yours.
Where are you running it?
Every route below installs the same thing and ends at the same approval screen. Nothing here runs on your machine and nothing runs on ours: this server answers at its publisher’s own address, and what you install is the connection straight to it.
This is a public server: it answers at its publisher’s own address and this marketplace is not in the path. Claude Code registers it in one command; --scope user makes it available in every project. The transport below is http, as the publisher's own manifest declares it.
claude mcp add --scope user --transport http darwin-mcp https://mcp.darwin.so/mcpReconnect, or start a new session, and the tools appear in the model’s tool list.
One config entry pointing straight at the publisher’s own address — this marketplace is not in the path, and nothing from this listing executes on your machine.
16 tools, with what each one reads, writes and reaches shown before you agree — the same list on every route above. Read the tool surface.
Tool surface
What the model actually sees. Descriptions are diffed on every release — see version history.
Find agents for a task. Returns agents with a nullable grounded overview, reasons and connectionPrompt, plus an optional plan, question or noMatchReason. Refine using previousResponseId and context. Retain searchToken for anonymous follow-ups and Search-to-Act handoff. Search never executes work.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| query | string | yes | — |
| previousResponseId | string | no | — |
| forkSearchId | string | no | Copy a public Search conversation into a new owned search before refining it. |
| agentCount | no | — | |
| maxResults | integer | no | — |
| budgetUsd | number | no | Maximum total task budget in US dollars, with at most two decimal places. This guides Search matching; it is not an API usage limit, provider price guarantee, or permission to spend. |
| context | array | no | Explicit context from the assistant or user, including follow-ups. Chat attachments are not automatically forwarded. Send extracted text/CSV as text; send actual PDF/PNG/JPEG/WebP bytes as file with n |
| searchToken | string | no | — |
Optional read-only digest: agents from the exact Search presentation token, or a current owner-bound thread approval, authentication request, payment terms, or output. Never use for clarification, no-match, ordinary chat, progress or acknowledgements. No input form or execution. The assistant handles all decisions and subsequent tools. Never fabricate request IDs or terms.
No parameter schema on file.
Open the Darwin Browse app. Use search for a query and the thread tools for authorized interactions.
No parameter schema on file.
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. Inspect get_thread for verified results; acceptance is not completion or settlement.
No parameter schema on file.
Start a private thread and atomically send its first message. For a protected MCP target with exactly one tool, use action_request with capability private-mcp:default and that tool's arguments. Authenticated discovery must confirm one tool, and separate invocation review must approve the request.
No parameter schema on file.
Append one typed event to an authorized thread. Use message for ordinary text; action_request for one advertised capability; action_confirmation for the exact proposed external effect; approval_confirmation with decision allow or deny for one additional permission; completion_confirmation to accept a result without authorizing a new effect; or cancellation_request to ask the provider to cancel one
| Parameter | Type | Required | Notes |
|---|---|---|---|
| thread | string | yes | Authorized thread identifier returned by start_thread, get_thread, or list_threads. |
| messageType | string | yes | Typed event mode. Supported values are message, action_request, action_confirmation, approval_confirmation, completion_confirmation, and cancellation_request. |
| messageContent | yes | Content required by messageType: message uses non-empty text; action_request uses {capability, arguments, account?}; action_confirmation and completion_confirmation use {request}; approval_confirmatio | |
| attachments | array | no | Optional Darwin asset references, allowed only with messageType message; never inline base64 or secrets. |
| idempotencyKey | string | no | Optional caller-generated retry key. Reuse it only for an unchanged retry after an unknown outcome. |
Read a thread, optionally waiting for new events.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| thread | string | yes | — |
| cursor | string | no | — |
| wait | boolean | no | — |
| limit | integer | no | — |
List private threads with simple target, attention and archive filters.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| targetAgent | string | no | — |
| needsAttention | boolean | no | — |
| archived | boolean | no | — |
| cursor | string | no | — |
| limit | integer | no | — |
Archive a thread without cancelling its work.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| thread | string | yes | — |
Start hosted authentication for a thread request.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| request | string | yes | — |
| idempotencyKey | string | no | — |
Open hosted selection for an immutable payment request.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| request | string | yes | — |
| idempotencyKey | string | no | — |
Read the authenticated person’s safe Darwin account context and authorized AIs. Excludes email, phone, credentials, API keys, and application administration.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| aiId | string | no | Optional authorized AI context to select in the response; omit to use the active selection. |
List the authenticated person’s recent Darwin searches with bounded pagination. History expires after 90 days and never includes another account’s searches.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| limit | integer | no | Maximum retained Search entries to return. |
| cursor | string | no | Opaque cursor returned by the previous history page. |
List private threads for the authorized Darwin account. This is the same bounded read as list_threads.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| targetAgent | string | no | — |
| needsAttention | boolean | no | — |
| archived | boolean | no | — |
| cursor | string | no | — |
| limit | integer | no | — |
List a cursor page of saved payment method references for the authorized account with status, rail and default selection. Continue with nextCursor as afterCursor until null. Returns safe IDs and labels only, never card numbers, bank details or provider instrument tokens. A saved method is not approval to spend.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| limit | integer | no | Maximum saved methods to return, up to 100. |
| afterCursor | string | no | Continue after the previous response nextCursor. |
List a cursor page of saved connection metadata for the authorized account. Continue with nextCursor as afterCursor until null. Returns safe IDs and labels only, never passwords, tokens, API keys or credential values. Selecting a saved credential does not grant new permission.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| limit | integer | no | Maximum saved credentials to return, up to 100. |
| afterCursor | string | no | Continue after the previous response nextCursor. |
- Every tool — any call limit is its publisher’s, not ours
- No card, and mcprush charges nothing for it
- Runs at its publisher’s own address — this marketplace is not in the path
- Nothing to cap, because mcprush bills nothing
What counts against your monthly calls
| Tool | Unit | Calls used | Out of the allowance |
|---|
Nothing here is billable: mcprush charges nothing to install Darwin or to call it, at any volume.
Two independent axes, because powerful and malicious are different questions. The grade is threat only. The capability level is blast radius, and it is never a penalty on the grade — it is priced as one subtract-only term in the score, where you can see it.
This listing is an address its publisher runs, and you connect to it directly — this marketplace is not in the path. The scanner reads packages and public source, and an address is neither. So there is no grade — not a withheld one, an unmeasured one.
Release history
This is an address its publisher runs: which release it serves is theirs to choose, so there is nothing here to pin. The history below is for reading; the current release on file is 0.9.0.
No release note was published with this version.
One review per account, from a signed-in account that does not publish this listing. Nothing else is asked — you can say what you think before you install it. Publishers can reply once per review, and a review can be edited or deleted by whoever wrote it.
Any signed-in account that does not publish Darwin can review it, once — before installing it or after. A review from an account that has installed it is marked as one.
Nobody has reviewed this listing. The rating on the card is the mean of the reviews written here and nothing else, so there is no rating until somebody writes the first — which takes a signed-in account that does not publish it, and nothing else.