tailscale-network
A distributed model context server for AI agents, leveraging Tailscale for secure networking.
claude mcp add --transport stdio cavanaughdesign-tailscale-network-mcp-server docker compose up -d \ --env TAILSCALE_AUTH_KEY="<your-tailcale-auth-key>"
How to use
This MCP server implements a distributed Tailscale-enabled model context network designed to securely share and synchronize context data across central, regional, and edge components. It proxies interactions between an AI agent simulator and multiple context servers via a zero-trust Tailnet, enabling real-time updates, versioning, and caching across environments. Use the provided agent simulator and API endpoints to create, manage, and query conversation contexts, while the central authority coordinates versioning and consistency across regional servers and edge caches.
How to install
Prerequisites:
- Docker and Docker Compose installed on your host
- A Tailscale account and an auth key for secure networking
- Optional: Node.js for local development if you want to run components individually
Setup steps:
- Clone the repository:
git clone https://github.com/yourusername/tailscale-model-context-server.git
cd tailscale-model-context-server
- Create a .env file (or set environment variables) with your Tailnet auth key placeholder:
# .env
TAILSCALE_AUTH_KEY=tskey-auth-xxxxxxxxxxxxxx
- Start the full system with Docker Compose (the docker-compose file orchestrates central, regional, edge caches, and agent simulator):
docker-compose up -d
- If you need to run components individually for local development, you can start specific services with Docker commands or npm scripts (if available in the repo). For example, to run the central server in isolation via Docker, you might use:
# example placeholder; adapt to your repo's actual service definitions
docker run -d --name central-server tailscale-model-context-server-central
- Access the AI Agent Simulator at:
http://localhost:8080
- To stop the system:
docker-compose down
Notes:
- Ensure your Tailnet is properly configured and the auth key is valid.
- If you modify environment variables, restart the affected services so changes take effect.
Additional notes
Tips and common issues:
- If components fail to discover peers, verify that the Tailnet is connected and that the Tailscale daemon is running inside each container.
- Check logs for authentication errors or network reachability between central, regional, and edge servers.
- Use environment variables to tune Tailnet behavior (for example, Tailnet version or debug flags) as supported by your deployment.
- When deploying to cloud environments, ensure security groups/firewall rules allow necessary Tailnet traffic across regions.
- For local development, the Node.js runtime can be used to simulate agents or contexts if you install and run those components separately from the Docker-based orchestration.
Related MCP Servers
zen
Selfhosted notes app. Single golang binary, notes stored as markdown within SQLite, full-text search, very low resource usage
MCP -Deepseek_R1
A Model Context Protocol (MCP) server implementation connecting Claude Desktop with DeepSeek's language models (R1/V3)
mcp-fhir
A Model Context Protocol implementation for FHIR
mcp
Inkdrop Model Context Protocol Server
mcp-appium-gestures
This is a Model Context Protocol (MCP) server providing resources and tools for Appium mobile gestures using Actions API..
dubco -npm
The (Unofficial) dubco-mcp-server enables AI assistants to manage Dub.co short links via the Model Context Protocol. It provides three MCP tools: create_link for generating new short URLs, update_link for modifying existing links, and delete_link for removing short links.