Quickstart
👋 Welcome!
Archestra.AI is an open source "AI Platform" for your organization, whether it's a small startup or a large enterprise, helping you build and deploy practical agents that safely operate on internal and external data.

Archestra is an open source platform built from multiple components:
- ChatGPT-like UI
- Agent Builder
- Powerful MCP orchestrator
- Unique deterministic security layer (this is important!) preventing agents from going rogue
...and a few others built for those considering Archestra for real-world production, like OTEL exporter, cost tracker, and optimizer.
Fun fact: The team behind Archestra.AI previously worked on Grafana OnCall.
Run it locally to try!
We've prepared a all-in-one Docker image for you. Once it's up, follow http://localhost:3000
docker pull archestra/platform:latest;
docker run -p 9000:9000 -p 3000:3000 \
-e ARCHESTRA_QUICKSTART=true \
-v /var/run/docker.sock:/var/run/docker.sock \
-v archestra-postgres-data:/var/lib/postgresql/data \
-v archestra-app-data:/app/data \
archestra/platform;
Build Your First Agent
- Go to MCP Registry, add new and search for
microsoft__playwright-mcp, install it. - Gaze at Archestra running the MCP server in Kubernetes cluster! (If you want to make sure:
kubectl get pods) - Go to Agents and create "Archestra Docs Reader Agent" with this system prompt:
You're using playwright to answer questions about Archestra based on https://archestra.ai/docs/ - While editing the agent, make sure to enable all
microsoft__playwright-mcptools for it. - Go to Settings -> LLM API Keys and add your preferred provider.
- Go to Chat, choose the "Archestra Docs Reader Agent" and ask "How could I deploy Archestra?"

- Enjoy Archestra navigating the website for you!
Connect to Your Agent via MCP Gateway
Archestra is not just a Chat UI, it's a very capable MCP Gateway. Let's connect our "Archestra Docs Reader Agent" to Claude Code to make sure we could trigger it as an MCP server!
- Go to MCP Gateways, create a new one.
- Make sure to set "Archestra Docs Reader Agent" as a sub-agent.
- Save changes and Copy MCP configuration.
- Add MCP Gateway to your Claude. Here is my archestra config (your auth key will be different!):
{
"mcpServers": {
"archestra": {
"url": "http://localhost:9000/v1/mcp/e1b0272c-3839-4575-a49d-aabb864d638d",
"headers": {
"Authorization": "Bearer archestra_119220a7bfc485d66b678d3e9fb2db36"
}
}
}
}
I converted it to the command for claude:
claude mcp add archestra "http://localhost:9000/v1/mcp/e1b0272c-3839-4575-a49d-aabb864d638d" --transport http --header "Authorization: Bearer archestra_119220a7bfc485d66b678d3e9fb2db36"
, and now ask Claude:
Ask archestra to give you all the deployment options.
Feel the magic ✨
What's next?
- Chat Interface - Interact with AI agents using MCP tools
- Agents - Build autonomous agents with no-code
- LLM Proxy - Secure your AI applications
- MCP Gateway - Connect external systems to agents
- Deployment - Deploy Archestra in production
- Brew By Weight? Brew By AI! - Fun post about building MCP, running it in Archestra to get the perfect espresso.
- Join our Slack Community - Connect with other users and the team