httptoolkit-mcp
An MCP server that enables AI assistants to control HTTP Toolkit for intercepting, inspecting, and debugging HTTP(S) traffic from browsers, mobile devices, and Docker containers. It provides tools for server management, interceptor activation, and sending HTTP requests through natural language commands.
README
httptoolkit-mcp
A Model Context Protocol (MCP) server that provides AI assistants with full control over HTTP Toolkit — the powerful, open-source HTTP(S) debugging proxy.
Intercept, inspect, and debug HTTP traffic from browsers, Android/iOS devices, Docker containers, JVM processes, and more — all through natural language via your AI assistant.
Features
- Live Traffic Capture — Capture intercepted HTTP traffic in real-time with full request/response headers and bodies
- Browser Interception — Launch Chrome or Firefox with traffic automatically routed through the proxy
- Android Interception — Intercept device traffic via ADB or target specific apps with Frida (bypasses certificate pinning)
- iOS Interception — Intercept specific iOS apps via Frida on jailbroken devices
- Docker Interception — Attach to running Docker containers and capture all outgoing HTTP traffic
- Terminal Interception — Open intercepted terminal sessions or configure existing ones
- JVM Attach — Attach to running Java/Kotlin/Clojure processes
- Electron Apps — Launch Electron applications with interception enabled
- HTTP Client — Send HTTP requests through the proxy with full control over method, headers, and body
- Server Management — Query configuration, manage interceptors, and control the server lifecycle
- Zero-Config with Desktop App — Automatically detects auth token from the running HTTP Toolkit desktop app
Prerequisites
- HTTP Toolkit installed and running (desktop app or server)
- Node.js >= 18
Installation
Using npx (recommended)
No installation required — just configure your MCP client:
{
"mcpServers": {
"httptoolkit": {
"command": "npx",
"args": ["-y", "httptoolkit-mcp"]
}
}
}
Global install
npm install -g httptoolkit-mcp
From source
git clone https://github.com/fdciabdul/httptoolkit-mcp.git
cd httptoolkit-mcp
npm install
npm run build
Configuration
Claude Code
Add to ~/.claude/settings.json or ~/.claude.json:
{
"mcpServers": {
"httptoolkit": {
"command": "npx",
"args": ["-y", "httptoolkit-mcp"]
}
}
}
Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"httptoolkit": {
"command": "npx",
"args": ["-y", "httptoolkit-mcp"]
}
}
}
Environment Variables
| Variable | Description | Default |
|---|---|---|
HTK_SERVER_URL |
HTTP Toolkit management API URL | http://127.0.0.1:45457 |
HTK_SERVER_TOKEN |
Auth token (auto-detected from desktop app) | Auto-detected |
HTK_ADMIN_URL |
Mockttp admin API URL | http://127.0.0.1:45456 |
Note: When using the HTTP Toolkit desktop app, the auth token is automatically detected from the running process — no manual configuration needed.
Example with manual environment variables:
{
"mcpServers": {
"httptoolkit": {
"command": "npx",
"args": ["-y", "httptoolkit-mcp"],
"env": {
"HTK_SERVER_URL": "http://127.0.0.1:45457",
"HTK_SERVER_TOKEN": "your-token-here"
}
}
}
}
Available Tools
Traffic Capture
| Tool | Description |
|---|---|
capture_traffic |
Capture live intercepted HTTP traffic with full request/response bodies. Subscribes to the active HTTP Toolkit session via WebSocket. Supports auto-detection or manual session ID. |
Server Management
| Tool | Description |
|---|---|
get_version |
Get the HTTP Toolkit server version |
get_config |
Get proxy configuration (certificates, DNS, system proxy) |
get_network_interfaces |
List all network interfaces |
trigger_update |
Trigger a server update check |
shutdown_server |
Shutdown the HTTP Toolkit server |
Interceptor Management
| Tool | Description |
|---|---|
list_interceptors |
List all available interceptors and their status |
get_interceptor_metadata |
Get detailed metadata for a specific interceptor |
activate_interceptor |
Generic interceptor activation with custom options |
deactivate_interceptor |
Deactivate a running interceptor |
Browser Interception
| Tool | Description |
|---|---|
intercept_chrome |
Launch a fresh Chrome window with interception |
intercept_firefox |
Launch a fresh Firefox window with interception |
Terminal Interception
| Tool | Description |
|---|---|
intercept_fresh_terminal |
Open a new terminal with interception enabled |
intercept_existing_terminal |
Get commands to enable interception in an existing terminal |
Docker Interception
| Tool | Description |
|---|---|
intercept_docker_container |
Attach to a running Docker container |
Android Interception
| Tool | Description |
|---|---|
intercept_android_adb |
Intercept an Android device/emulator via ADB |
frida_android_setup |
Set up Frida on an Android device |
frida_android_launch |
Launch Frida server on an Android device |
frida_android_intercept |
Intercept a specific Android app via Frida |
iOS Interception
| Tool | Description |
|---|---|
frida_ios_intercept |
Intercept a specific iOS app via Frida |
Application Interception
| Tool | Description |
|---|---|
intercept_jvm |
Attach to a running JVM process |
intercept_electron |
Launch an Electron app with interception |
HTTP Client
| Tool | Description |
|---|---|
send_http_request |
Send an HTTP request through the proxy |
Usage Examples
Once configured, you can ask your AI assistant things like:
- "Capture the HTTP traffic from my intercepted Chrome for 10 seconds"
- "List all available interceptors"
- "Intercept Chrome on port 8000"
- "Show me the connected Android devices"
- "Attach to the Docker container running my API"
- "Send a GET request to https://api.example.com/users"
- "Set up Frida on my Android device and intercept the target app"
- "Open an intercepted terminal session"
Capture Traffic Example
User: "Capture traffic from my browser for 5 seconds"
capture_traffic({ duration: 5, sessionId: "2474b580-482e-4a79-8488-121583d466e1" })
Result:
{
"capturedExchanges": 2,
"exchanges": [
{
"request": {
"method": "GET",
"url": "https://api.example.com/users",
"headers": { "host": "api.example.com", ... },
},
"response": {
"statusCode": 200,
"headers": { "content-type": "application/json", ... },
"body": "[{\"id\": 1, \"name\": \"John\"}]"
}
}
]
}
Architecture
┌─────────────────┐ stdio ┌──────────────────┐ REST API ┌─────────────────────┐
│ AI Assistant │◄─────────────►│ httptoolkit-mcp │◄────────────►│ httptoolkit-server │
│ (Claude, etc.) │ MCP │ (this project) │ :45457 │ (HTTP Toolkit) │
└─────────────────┘ └──────────────────┘ └─────────────────────┘
│ │
│ WebSocket ┌────────┴────────┐
│ (traffic capture) │ Mockttp Proxy │
└──────────────────────────►│ :45456 (admin) │
│ :8000 (proxy) │
└────────┬────────┘
│
┌───────────────┼───────────────┐
│ │ │
Browsers Android/iOS Docker
Devices Containers
Credits
This project is an MCP interface for HTTP Toolkit, created by Tim Perry (@pimterry).
- HTTP Toolkit — github.com/httptoolkit
- HTTP Toolkit Server — github.com/httptoolkit/httptoolkit-server
- Tim Perry — github.com/pimterry — Creator & maintainer of HTTP Toolkit
HTTP Toolkit is a beautiful, open-source tool for debugging, testing, and building with HTTP(S). If you find it useful, consider supporting the project.
License
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.