> For the complete documentation index, see [llms.txt](https://docs.oncompute.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.oncompute.ai/on-mcp/quickstart.md).

# Overview

**ON MCP** is a [Model Context Protocol](https://modelcontextprotocol.io) server that bridges AI agents to the **Ocean Network**. Connect it to Claude, Cursor, ChatGPT, Gemini or VS Code / GitHub Copilot, and the agent can discover compute providers, write and validate C2D algorithms, run free or paid jobs, and manage persistent storage all in natural language.

It exposes roughly **80 MCP tools** plus searchable documentation **resources**, mapping the full surface of Ocean's `ocean-node` and `libp2p` capabilities into something an agent can use directly.

#### How it works

The server is **P2P-first**: on startup it joins the Ocean libp2p network, so it discovers and talks to nodes directly over the DHT rather than depending on a single HTTP gateway. When your agent asks for compute, it resolves a node, negotiates an environment, runs the job near the data, and pulls back only the results.

#### The three flows

Once connected, the agent can work along three flows:

* **Demand side: run a job.** Find a provider, write an algorithm, and run free or paid compute against a URL or a private file.
* **Supply side: publish data.** Make datasets and algorithms available on the network.
* **Node operation.** Inspect node status, environments, and jobs.

#### Your keys stay with you

The server **never holds your funds and never signs anything**. On-chain operations (paying for compute, escrow deposits) return **unsigned transactions,** you authorize and broadcast them yourself, e.g. on the Ocean Network Dashboard. For quick, free experiments the server can use an ephemeral key, so you can run real compute with no wallet at all.

{% hint style="info" %}
**How does this compare to the other tools?** The Ocean Orchestrator is an interactive VS Code extension for authoring and launching jobs by hand. The Ocean CLI drives the network from your terminal. **ON Compute MCP** is the *agent* interface, you describe the goal and the AI orchestrates the tools for you.
{% endhint %}

#### Next steps

* [Connect & Configure — get the server connected to your AI client.](/on-mcp/publish-your-docs.md)
* [Walkthrough: From Connect to Paid Compute — a full end-to-end run.](/on-mcp/walkthrough-from-connect-to-paid-compute.md)
* [Tools & Resources — the capabilities reference.](/on-mcp/tools-and-resources.md)


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.oncompute.ai/on-mcp/quickstart.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.