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

# MCP Server

The **Doma MCP server** is a hosted [Model Context Protocol](https://modelcontextprotocol.io) server that gives any MCP-capable AI agent direct access to Doma: domain lookup and search, availability and pricing, DNS, marketplace listings and quotes, token prices and wallet balances — and, once authorized, buying listed domains, swapping tokens, bridging assets to other chains, and resting limit orders from the user's own Doma wallet.

One URL in your MCP client and the agent discovers the tools itself — **no API key, no contract ABIs, no Seaport or orderbook code**.

## Endpoints

| Network      | URL                                | Chain ID |
| ------------ | ---------------------------------- | -------- |
| Doma Mainnet | `https://mcp.doma.xyz/mcp`         | `97477`  |
| Doma Testnet | `https://mcp-testnet.doma.xyz/mcp` | `97476`  |

Both speak MCP over **Streamable HTTP**. Each server executes on exactly one network — `server.ping.v1` reports which. To use both, connect both under distinct names (`doma`, `doma-testnet`).

{% hint style="info" %}
Pasting the bare origin (`https://mcp.doma.xyz`) also works — the server routes protocol traffic from `/` to `/mcp`. The `/mcp` path is the canonical one to configure.
{% endhint %}

## Two tiers

* **Read tools — no authentication.** Lookups, search, pricing, DNS, listings, offers, quotes, token data, balances. Connect and start asking; nothing is signed, nothing is spent.
* **Execute tools — OAuth authorization required.** Buying a listing, swapping tokens, and bridging assets to other chains sign and broadcast real transactions from the user's Doma wallet, gated by an OAuth 2.1 flow, a **USD spend allowance the user sets**, a per-transaction ceiling, and a rate limit. Limit orders are authorized the same way, but — because they fill off-session — are bounded by a standing on-chain router allowance rather than the USD budget. See [Authorization & Spend Controls](/agentic-commerce/mcp-server/authorization.md).

Execute tools are still *listed* to unauthorized clients, so an agent can see what's possible; calling one without a token returns a standard `401` challenge, which is the cue for your client to run the OAuth flow.

Execution on Doma chains is **gasless** — the user needs no ETH on the Doma chain to trade.

## Which surface should I use?

Doma exposes agent-facing capability three ways. They overlap deliberately; pick by how much control you want.

| Surface                                                | Best for                                                                                                                             | Setup                  |
| ------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------ | ---------------------- |
| **MCP server** (this section)                          | Any MCP client — Claude, Cursor, Codex, your own agent. Tools are discovered automatically and stay current with no action from you. | One URL                |
| [**AgentRoot skills**](/agentic-commerce/agentroot.md) | Agents that load `SKILL.md` files and want Doma's own instructions for multi-step workflows, discovered over DNS.                    | One-time skill install |
| [**Doma CLI**](/agentic-commerce/doma-cli.md)          | Humans in a terminal, scripts, and CI — plus operations the MCP server doesn't expose yet (bridging, subdomains, selling).           | `npm install -g`       |

If you are wiring up an agent and don't have a reason to prefer another path, start here.

## Where to go next

* [**Connect Your Client**](/agentic-commerce/mcp-server/connect.md) — copy-paste config for Claude Code, Claude, Cursor, VS Code, Codex, Windsurf, and anything else.
* [**Authorization & Spend Controls**](/agentic-commerce/mcp-server/authorization.md) — how the OAuth flow and the spend allowance work, and how to revoke.
* [**Tool Reference**](/agentic-commerce/mcp-server/tools.md) — the full tool surface at a glance, and what isn't exposed yet.
* [**Examples**](/agentic-commerce/mcp-server/examples.md) — real prompts and what the agent does with them.


---

# 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 by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.doma.xyz/agentic-commerce/mcp-server.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

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.