Sandbox
@soth-ai/mcp-reticle

MCP traffic proxy and inspector for agents

Reticle sits between your agent and MCP servers so you can inspect JSON-RPC messages as they happen. It can wrap stdio servers, proxy HTTP-based transports, launch a desktop UI, or run headless for logging and telemetry.

118 stars46 forksRustUpdated 8mo ago
Who it's for

People who build with MCP-connected agents and want to inspect tool traffic, latency, and failures.

What it delivers

You can debug MCP integrations by seeing the exact messages, timing, and errors your agent exchanges with servers.

What it does

Real-time message inspection

Shows raw JSON-RPC requests, notifications, and responses as they pass through the proxy.

Request and response correlation

Matches each request with its response so you can follow one interaction end to end.

Latency and token profiling

Measures timing and estimates token usage for MCP calls.

Crash and stderr capture

Records server stderr output and crashes alongside the traffic.

Session recording and export

Lets you record sessions and export logs for later review.

Multiple transport support

Works with stdio, Streamable HTTP, WebSocket, and HTTP/SSE transports.

CLI and desktop UI

Provides `mcp-reticle run`, `proxy`, `ui`, and `daemon` commands for different debugging setups.

How to get it

  1. 1Run
    # npm
    npm install -g mcp-reticle
    
    # pip
    pip install mcp-reticle
    
    # Homebrew
    brew install labterminal/mcp-reticle/mcp-reticle
  2. 2Run
    mcp-reticle ui
  3. 3Run
    mcp-reticle run --log -- npx -y @modelcontextprotocol/server-memory
  4. 4Run
    mcp-reticle proxy --name api --upstream http://localhost:8080 --listen 3001
  5. 5If you prefer building from source
    git clone https://github.com/labterminal/mcp-reticle.git
    cd mcp-reticle
    just build

README

Reticle LogoReticle Logo

RETICLE

The Wireshark for the Model Context Protocol (MCP)

License: MIT npm PyPI Docs (Wiki) Platform

See what your agent sees.

Reticle intercepts, visualizes, and profiles MCP JSON-RPC traffic in real time — designed for microsecond-level overhead.

Quick start · Install · Docs · Security · Contributing

Reticle Screenshot


What is Reticle?

Reticle is a proxy + UI for debugging MCP integrations:

  • inspect raw JSON-RPC messages (requests / notifications / responses)
  • correlate request ↔ response instantly
  • profile latency and token estimates
  • capture server stderr and crashes
  • record sessions and export logs

Supported transports: stdio, Streamable HTTP, WebSocket, HTTP/SSE.


Quick start

1) Install

# npm
npm install -g mcp-reticle

# pip
pip install mcp-reticle

# Homebrew
brew install labterminal/mcp-reticle/mcp-reticle

2) Wrap your MCP server (stdio)

Replace your MCP server command with mcp-reticle run --name <name> -- <command...>.

Example (Claude Desktop-style config):

{
  "mcpServers": {
    "filesystem": {
      "command": "mcp-reticle",
      "args": ["run", "--name", "filesystem", "--", "npx", "-y", "@modelcontextprotocol/server-filesystem", "/Users/me/work"]
    }
  }
}

3) Launch the UI

mcp-reticle ui

Optional: log-only mode (no UI)

mcp-reticle run --log -- npx -y @modelcontextprotocol/server-memory

Optional: proxy an HTTP-based MCP server

mcp-reticle proxy --name api --upstream http://localhost:8080 --listen 3001

Installation

If you prefer building from source:

git clone https://github.com/labterminal/mcp-reticle.git
cd mcp-reticle
just build

Documentation

All guides and deep dives are in the GitHub Wiki:


CLI overview

CommandPurpose
mcp-reticle runWrap stdio MCP servers and inspect traffic
mcp-reticle proxyReverse proxy HTTP/SSE/WebSocket transports
mcp-reticle uiLaunch the desktop UI
mcp-reticle daemonHeadless telemetry hub

Full details: CLI reference


Security & privacy

Reticle can capture tool inputs/outputs and server stderr. Treat recordings and exports as sensitive artifacts.

Recommended reading: Security & privacy


Contributing


License

MIT — see LICENSE.

Files in the repo

Repository payload18 top-level entries
  • .github
  • crates
  • Formula
  • frontend
  • packages
  • scripts
  • src-tauri
  • .editorconfig
  • .gitignore
  • CHANGELOG.md
  • CODE_OF_CONDUCT.md
  • CONTRIBUTING.md
  • justfile
  • LICENSE
  • README.md
  • rust-toolchain.toml
  • SECURITY.md
  • TESTING.md

Discussion (0)

Ask about usage, or say what you built with it

Sign in to join the discussion.

No comments yet. Be the first to say what this is good for.

More tools

JuliusBrussee/
caveman

🪨 why use many token when few token do trick — Claude Code skill that cuts 65% of tokens by talking like caveman

105k
1 add
MemPalace/
mempalace

The best-benchmarked open-source AI memory system. And it's free.

59k
stablyai/
orca

Orca is the ADE for working with a fleet of parallel agents. Run any coding agent with your own subscription. Available on desktop, mobile and remote runtime.

66k

A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Grok Build & Hermes Agent. Only official website: ccswitch.io

132k

Never stop coding. Free MIT AI gateway: one endpoint, 352 providers (150+ free), 1200+ models Kimi, Claude, GPT, Gemini, GLM, DeepSeek, MiniMax. Works with Claude Code, Codex, Cursor, OpenCode, Cline & Copilot. Quota-aware auto-fallback, RTK+Caveman compression saves 15-95% tokens, MCP/A2A, Desktop/PWA. Built by 550+ contributors

64k
headroomlabs-ai/
headroom

Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.

71k