High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.
MCP server and CLI for MSW mocks
This repo bridges AI agents and Mock Service Worker so they can manage network mocks from the terminal or through MCP tools. It uses a browser client plus a core session daemon to push handler changes live, with no page reload or app restart.
Builders who want their agent to add and edit MSW mocks while they build frontend features.
You can change network mocks from an agent or CLI and see them take effect immediately.
What it does
Dynamic mock updates
Add, update, and remove MSW handlers without restarting the app.
MCP tool access
Expose mock management through `msw-mcp` for MCP clients.
CLI control
Use `msw-cli` to open a session, add handlers, check status, and close it.
Browser bridge
Send handler updates through a WebSocket bridge to the local MSW client.
Setup scaffolding
Create the MSW and client wiring with `msw-cli setup` or the `/msw-setup` prompt.
Agent skills
Provide `skills/msw-cli` and `skills/msw-setup` for agent workflows.
How to get it
- 1Start a session, then drive handlers from the shell
msw-cli open msw-cli add "http.get('/api/user', () => HttpResponse.json({ id: 1 }))" msw-cli status msw-cli close - 2Install dependencies
pnpm install
- 3Build the project
pnpm build
- 4Use the CLI: You can navigate to the CLI directory or use it directly via node after…
node packages/msw-cli/build/index.js help
README
MSW AI Agent Bridge Monorepo
This monorepo provides a suite of tools that bridge AI agents with Mock Service Worker (MSW). It enables dynamic addition, updating, and removal of MSW network mocks directly from a command-line interface or AI tool execution, completely independently of the application source code restarts.
Demo
Walkthrough of the MCP-driven MSW workflow (browser bridge, live handler updates):
https://github.com/user-attachments/assets/06959759-b198-4cd1-ae2b-11ede32acd8e
Key features
- 🤖 Generate MSW handlers with AI
- ⚡ Update mocks without reloading the page
- 🌉 WebSocket bridge between Cli/MCP and the browser
- 🛠️ Scaffold MSW setup with
msw-cli setupor MCP's/msw-setupprompt - 💻 Terminal and script control via
msw-cli - ⚙️ Configurable port, persistence, and client targeting
- ✨ Agent skills for AI agents to use
Structure
This project is set up as a monorepo using pnpm and turbo.
apps/docs(@msw-mcp/docs): Starlight documentation site for MSW Agent Tools (CLI, MCP, guides). Source lives underapps/docs/src/content/docs/.packages/msw-cli: The command-line interface (msw-cli) used by AI agents or developers to send dynamic handler updates. Runmsw-cli setupto scaffold a project, thenmsw-cli openplus handler commands.packages/msw-mcp: The published Model Context Protocol server (msw-mcp) so assistants can manage mocks via MCP tools.packages/client(@msw-mcp/client): The browser/environment client that listens for updates and registers them with the local MSW setup.packages/core(@msw-mcp/core): The core daemon process and session management logic that brokers communication between the CLI or MCP server and the client.
Documentation site
From the repo root:
pnpm --filter @msw-mcp/docs dev
Opens the local docs dev server (default Astro port, often http://localhost:4321). Production search (Pagefind) is easiest to verify after pnpm --filter @msw-mcp/docs build and pnpm --filter @msw-mcp/docs preview.
More detail: apps/docs/README.md.
Quick setup: msw-cli
-
Install the CLI globally:
npm install -g msw-cli(or use your package manager’s equivalent). -
In a new project, run
msw-cli setupand follow the printed steps for MSW +@msw-mcp/client. -
Start a session, then drive handlers from the shell:
msw-cli open msw-cli add "http.get('/api/user', () => HttpResponse.json({ id: 1 }))" msw-cli status msw-cli close
Full command reference, flags, and workflows: packages/msw-cli/README.md.
Quick setup: msw-mcp (MCP)
-
Register the server in your MCP client (example for JSON config):
{ "mcpServers": { "msw-mcp": { "command": "npx", "args": ["msw-mcp@latest"] } } } -
In the client, use the
/msw-setupprompt (or follow manual steps) so the browser app loadsmsw,@msw-mcp/client, and the worker bridge.
Installation options (VS Code, Claude Code, Cursor, etc.), architecture, MCP tools, and manual setup: packages/msw-mcp/README.md.
Development Scripts
Run these scripts from the root directory using pnpm:
pnpm build: Compiles all packages using Turborepo.pnpm dev: Starts the development pipelines for all packages.pnpm format: Formats the code using Prettier.pnpm lint: Runs linting checks across the repository.
Getting Started
-
Install dependencies:
pnpm install -
Build the project:
pnpm build -
Use the CLI:
You can navigate to the CLI directory or use it directly via node after building:node packages/msw-cli/build/index.js help
License
This project is licensed under the MIT License — see LICENSE.
Files in the repo
- .cursor
- .github
- .husky
- apps
- packages
- skills
- .gitignore
- .nvmrc
- .prettierrc
- AGENTS.md
- CLAUDE.md
- LICENSE
- package.json
- pnpm-lock.yaml
- pnpm-workspace.yaml
- README.md
- turbo.json
Discussion (0)
Ask about usage, or say what you built with itSign in to join the discussion.
No comments yet. Be the first to say what this is good for.
More connectors

Universal provider proxy for OpenAI Codex & Claude Code — use any LLM (Claude, Gemini, Grok, DeepSeek, Ollama…) with Codex CLI, App, SDK, and Claude Code
Git-native persistent memory for AI coding agents. Implements Google OKF v0.2 with sub-300µs in-memory BM25 search, embedded MCP server, and progressive disclosure. Slashes token bloat by 80% with zero external databases or dependencies. Built in pure Go.
Local-first code intelligence graph for MCP and CLI. Builds a persistent map of your codebase so AI coding tools read only what matters, with benchmarked context reductions on reviews and large-repo workflows.
Stop your AI from making things up — it proposes, deterministic tools decide, every claim checked against ground truth with evidence. Grounded facts and context survive resets. Reverse engineering is the proving ground. MCP server + CLI.
20 MB lightweight cross-platform database client for 90+ databases, including MySQL, PostgreSQL, SQLite, Redis, MongoDB, DuckDB, SQL Server, and Dameng. Built-in AI, MCP Server, CLI, desktop and Docker. | 轻量级跨平台数据库管理工具,支持 MySQL、PostgreSQL、SQLite、Redis、MongoDB、达梦等 90+ 数据库,提供桌面端、Docker、CLI、内置 AI 助手和 MCP Server。