Add an MCP Server to Your Stack in One Line
Connect your AI coding agent to real infrastructure. Add the Open Source Cloud MCP server in one line and let your agent provision and run open-source services.
If you want your AI coding agent to do more than write code, give it an MCP server that connects to real infrastructure. On Open Source Cloud you add one with a single command, authenticate once in the browser, and from then on your agent can provision a database, an object store, a search index, or any of 180+ open-source services and read back their live connection details, all from inside the chat. No glue code, no separate dashboard, no runtime lock-in. Here is what that gives you and how to set it up.
What an MCP server actually does for an agent
The Model Context Protocol is the standard way an AI agent talks to outside tools and data. On its own, a coding agent can only generate text. With an MCP server attached, it can take actions: call an API, query a system, or in this case, stand up and manage real services. The protocol is the wiring; the question is what is on the other end of it. What is on the other end of the OSC MCP server is your actual infrastructure. Not a read-only description of it, not a sandbox, but the real catalog of open-source services and your own running instances. That is the difference between an agent that can tell you how to deploy Postgres and an agent that just deployed one and handed you the connection string.
One line to connect it
Add the OSC MCP server to your AI coding tool with a single command: "claude mcp add --transport http osc https://mcp.osaas.io/mcp". The first time the tool calls it, OSC authenticates you in the browser using OAuth and returns a token automatically. There is nothing to hand-edit and no token to copy and paste. The same one-line shape works for the other MCP-capable coding tools too; each one handles the browser auth for you. Once connected, your agent gains tens of tools plus a live view of your own services. It can list what you already have running, provision something new, fetch credentials, and check health, the same operations you would do by hand in the dashboard, now available to the agent mid-task.
From "suggest a stack" to "run the stack"
The practical effect is that the deploy step folds into the build step. You ask your agent to add a cache, and it provisions one and wires the connection string into your code. You ask it to add full-text search, and the search service is running before you finish the sentence. The services your agent reaches for are real open-source projects from the catalog, 180+ of them across 8 categories, so you are assembling production infrastructure, not stubbing a demo.
Open by default, so the exit is built in
Because every service the MCP server can provision is unmodified open source, none of this is a one-way door. Each piece is replaceable on its own, and the same services run on a hyperscaler account, your own Kubernetes cluster, or your own hardware. If you ever want to move a service off the platform, that is usually a connection-string change, not a rewrite. The convenience of provisioning through your agent does not cost you the ability to leave. Connect the MCP server once, and your AI agent stops being a code generator that hands work back to you and starts being one that can finish the deploy itself.
Frequently Asked Questions
How do I add an MCP server to Claude Code or another AI coding tool?
Run one command. For Open Source Cloud it is "claude mcp add --transport http osc https://mcp.osaas.io/mcp". The first call authenticates you in the browser over OAuth and stores the token automatically, with no config file to edit and no token to paste. Other MCP-capable tools use the same one-line shape and handle the browser auth for you.
What can my agent do once the OSC MCP server is connected?
It gains tens of tools plus a live view of your own running services. It can list, provision, and configure open-source services from a deep catalog across 8 categories, fetch live connection details and credentials, and check instance health, all from inside your coding session rather than a separate dashboard.
Is using a managed MCP server a lock-in risk?
No, when the services behind it are open source. On Open Source Cloud every service the MCP can provision is unmodified open source and individually replaceable. The same stack runs on a hyperscaler, your own Kubernetes cluster, or on-premise, and moving a service is typically a connection-string change rather than an application rewrite.
Related Posts
Deploy What Your AI Just Built
Your AI coding assistant wrote the app. Here is how to deploy the backend on Open Source Cloud in one line, with no DevOps and no runtime lock-in.
You Don't Buy a Streaming Backend Anymore. You Provision One.
A streaming backend used to be a buy-or-build decision. With dozens of open-source media services on Open Source Cloud, you provision one in an afternoon and own every piece.
Open-Source Services You Can Deploy Today, by Category
A tour of the open-source services you can provision on Open Source Cloud, grouped by category, all unmodified open source and yours to move anytime.