Shift MCP Server
A lightweight coordination layer for multiple AI agents working on the same codebase, providing check-in and check-out tools via STDIO or Streamable HTTP.
README
<div align="center"> <h1>@cyanheads/shift-mcp-server</h1> <p><b>Lightweight coordination layer for multiple AI agents working on the same codebase. STDIO & Streamable HTTP</b></p> <p><b>2 Tools · 1 Resource</b></p> </div>
<div align="center">
</div>
Tools
Two tools for multi-agent coordination:
| Tool Name | Description |
|---|---|
shift_check_in |
Register or update a worker session. Returns worker ID, coordination instructions, and active peers. |
shift_check_out |
End a working session. Removes from active worker list. |
shift_check_in
Register a new worker or update an existing session. Called at the start of every working session.
- Accepts a gist of current work and optional file paths being modified
- Returns a 6-character worker ID, coordination protocol, and the full active workers table
- Pass an existing worker ID to update your session (patch semantics — omitted fields preserved)
- Error responses include the active workers table so agents can self-identify or start fresh
shift_check_out
End a working session and remove from the active worker list.
- Accepts a worker ID and optional summary of what was accomplished
- Idempotent — succeeds silently if the worker ID doesn't exist or was already checked out
Resources
| URI | Description |
|---|---|
shift://status |
All currently active workers with gists, declared files, and timestamps. |
Features
Built on @cyanheads/mcp-ts-core:
- Declarative tool definitions — single file per tool, framework handles registration and validation
- Unified error handling across all tools
- Structured logging with request-scoped context
- Runs locally (stdio/HTTP) from the same codebase
Coordination-specific:
- In-memory worker session store — no database, no filesystem writes, clears on restart
- Coordination protocol injected on every check-in so agents know how to behave
- Active workers table returned with every response for situational awareness
- Patch semantics on session updates — only provided fields change
Getting Started
MCP Client Config
Add to your MCP client config (e.g., claude_desktop_config.json):
{
"mcpServers": {
"shift": {
"type": "stdio",
"command": "bunx",
"args": ["@cyanheads/shift-mcp-server@latest"],
"env": {
"MCP_TRANSPORT_TYPE": "stdio"
}
}
}
}
Prerequisites
- Bun v1.2.0 or higher
Installation
git clone https://github.com/cyanheads/shift-mcp-server.git
cd shift-mcp-server
bun install
Configuration
No server-specific environment variables required. Framework defaults:
| Variable | Description | Default |
|---|---|---|
MCP_TRANSPORT_TYPE |
Transport: stdio or http. |
stdio |
MCP_HTTP_PORT |
Port for HTTP server. | 3010 |
MCP_HTTP_HOST |
Hostname for HTTP server. | 127.0.0.1 |
MCP_LOG_LEVEL |
Log level (RFC 5424). | info |
Running the Server
Local Development
bun run build
bun run start:stdio # or start:http
Dev mode with watch:
bun run dev:stdio # or dev:http
Checks and tests:
bun run devcheck # Lints, formats, type-checks
bun test # Runs test suite
Docker
docker build -t shift-mcp-server .
docker run -p 3010:3010 shift-mcp-server
Project Structure
| Directory | Purpose |
|---|---|
src/index.ts |
Entry point — registers tools and resources with createApp(). |
src/mcp-server/tools/definitions/ |
Tool definitions (check-in.tool.ts, check-out.tool.ts). |
src/mcp-server/tools/definitions/worker-store.ts |
In-memory worker session store and formatting utilities. |
src/mcp-server/resources/definitions/ |
Resource definitions (status.resource.ts). |
Development Guide
See CLAUDE.md for development guidelines and architectural rules. The short version:
- Handlers throw, framework catches — no
try/catchin tool logic - Use
ctx.logfor request-scoped logging - Register new tools and resources in
src/index.ts
Contributing
Issues and pull requests are welcome. Run checks before submitting:
bun run devcheck
bun test
License
Apache-2.0 — see LICENSE for details.
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
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
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.