Connect Dock to any agent.

54 setup guides, one per client. Every agent that speaks MCP or plain HTTPS can call Dock's 37 tools. Pick yours below for exact config paths, verification commands, and troubleshooting.

IDE-native coding agents

MCP-first coding agents that live in your editor. Drop a config block and reload.

Claude Code

Terminal. OAuth or Bearer. ~/.claude.json.

→

Cursor

Project or user scope. Remote MCP over HTTP.

→

Windsurf

Cascade agent. ~/.codeium/windsurf/mcp_config.json.

→

Zed

Context servers in settings.json.

→

Cline

VS Code extension. Side-panel config.

→

Continue

~/.continue/config.json. Cross-IDE.

→

Aider

Terminal pair programmer. YAML config.

→

GitHub Copilot

Copilot Business/Enterprise agent mode.

→

JetBrains AI Assistant

IntelliJ, PyCharm, WebStorm.

→

Amazon Q Developer

AWS coding agent. ~/.aws/amazonq/mcp.json.

→

Kiro

AWS agentic IDE. ~/.kiro/mcp.json.

→

Sourcegraph Cody

Code-aware AI. Amp + Cody configs.

→

Tabnine

Enterprise chat + MCP.

→

Supermaven

Fast inline + chat MCP.

→

Desktop + chat clients

Conversational clients. Most use OAuth 2.1 and DCR, no token ceremony needed.

Claude Desktop

Custom connector in Settings. OAuth.

→

Claude.ai (web + Projects)

Browser custom connector. Account-level.

→

Claude Cowork

Sandboxed desktop agent. Remote MCP only.

→

ChatGPT (Custom GPT + Actions)

REST via Actions. OpenAPI schema.

→

Gemini CLI

Google's CLI. ~/.gemini/settings.json.

→

OpenAI Codex CLI

OpenAI's terminal agent. TOML config.

→

OpenClaw

Open-source Claude-Code-like. openclaw.json.

→

Warp

AI-first terminal. Agent Mode MCP.

→

Raycast AI

macOS launcher. Script Commands + secrets.

→

Self-hosted chat UIs

Run your own chat with any model. All accept remote MCP servers.

LibreChat

Self-hosted UI. librechat.yaml.

→

Open WebUI

ChatGPT-style UI. Admin tools panel.

→

AnythingLLM

Desktop + Docker. Agent Skills.

→

Ollama

Local models + ollama-mcp-bridge.

→

SDKs + agent frameworks

Build custom agents and integrations.

LangChain

langchain-mcp-adapters. Python + JS.

→

LangGraph

Stateful graphs. Uses LangChain adapters.

→

Mastra

TS-first. Built-in MCPClient.

→

CrewAI

Multi-agent Python. MCPServerAdapter.

→

AutoGen

Microsoft multi-agent. autogen-ext[mcp].

→

Semantic Kernel

Microsoft .NET + Python.

→

Pydantic AI

Typed agent framework.

→

Agno (Phidata)

Python multi-agent. MCPTools.

→

LlamaIndex

Data framework. llama-index-tools-mcp.

→

Vercel AI SDK

TS SDK. experimental_createMCPClient.

→

OpenAI Agents SDK

MCPServer helper.

→

Anthropic SDK

Built-in MCP Connector beta.

→

MCP TypeScript SDK

Raw MCP client for Node.

→

MCP Python SDK

Raw MCP client for Python.

→

Visual flow builders

Low-code canvases for LLM apps. All have MCP nodes.

Dify

LLM app platform. MCP tool in Studio.

→

Langflow

Visual LangChain. MCP Server node.

→

Flowise

Low-code LangChain. MCP Tool node.

→

Automation + no-code

Connect Dock to your existing workflows.

n8n

MCP Client Tool node + HTTP Request.

→

Zapier

Webhooks in, HTTP actions out.

→

App builders

AI app generators. REST integration via env secret + fetch.

Bolt.new

StackBlitz app builder.

→

v0

Vercel UI generator.

→

Lovable

Supabase-first app builder.

→

Replit Agent

Secret + fetch pattern.

→

Devin

Machine MCP servers + Knowledge.

→

Raw HTTP + scripts

Plain shell or language clients, no agent framework needed.

curl / bash

One-off ops, CI, quick tests.

→

Python

requests, httpx, or MCP SDK.

→

Node.js

Native fetch, EventSource, MCP SDK.

→