MCP server

Ask your AI assistant about your deployments

The DeployHQ MCP server address is below. It's meant for AI assistants, not for browsers. Add it to your assistant, and you can ask about your projects, servers and deployments in plain language. Nothing to install: add the address, sign in to DeployHQ, approve.

https://mcp.deployhq.com/mcp

Read the full documentation No account yet? Try DeployHQ free

What you can ask

Replace PROJECT with the name of one of your projects.

Which of my projects has gone the longest without a deployment?
Have any deployments of PROJECT failed recently?
Which servers does PROJECT deploy to, and which of them auto-deploy?
Which branch and revisions did the latest deployment of PROJECT use?
How long did the latest deployment of PROJECT take, and which servers did it go to?
Does PROJECT deploy automatically, and from which repository branch?

Connect your assistant

Claude Code

Run this in your terminal. Then type /mcp in Claude Code and sign in to DeployHQ.

Terminal

claude mcp add --transport http --scope user deployhq https://mcp.deployhq.com/mcp

Codex

Run this in your terminal. Codex then asks you to sign in to DeployHQ.

Terminal

codex mcp add deployhq --url https://mcp.deployhq.com/mcp

Gemini CLI

Run this in your terminal. Then type /mcp auth deployhq in Gemini CLI to sign in to DeployHQ.

Terminal

gemini mcp add --transport http deployhq https://mcp.deployhq.com/mcp

Cursor

Add this to ~/.cursor/mcp.json (or .cursor/mcp.json in a project), then choose Connect next to DeployHQ in Cursor's MCP settings and sign in.

mcp.json

{
  "mcpServers": {
    "deployhq": {
      "url": "https://mcp.deployhq.com/mcp"
    }
  }
}

VS Code

Add this to .vscode/mcp.json in your workspace, then choose Start above the server entry and sign in to DeployHQ.

mcp.json

{
  "servers": {
    "deployhq": {
      "type": "http",
      "url": "https://mcp.deployhq.com/mcp"
    }
  }
}

Claude

In Claude.ai, open Settings, then Connectors, then "Add custom connector".
Enter DeployHQ as the name and paste the address above as the URL, then add it.
Choose Connect, sign in to DeployHQ and approve the connection.

On Team and Enterprise plans, an organization owner adds the connector first, then each member connects individually.

ChatGPT

In ChatGPT, open Plugins, choose the plus button, then "Add custom MCP server".
Enter DeployHQ as the name and https://mcp.deployhq.com/mcp as the endpoint.
Choose OAuth for authentication, review the risk notice and continue.
Sign in to DeployHQ and approve the connection.

Which ChatGPT plans can add custom MCP servers, and the exact menu labels, are set by OpenAI and change from time to time. See OpenAI's help article.

What it can and cannot see

It can read

Accounts you can use (list_accounts)
Your projects (list_projects)
One project: whether automatic deployments are on, the repository type, address and branch (get_project)
The servers of a project: protocol, environment, branch, server group and auto-deploy (list_servers)
The deployments of a project, newest first, filterable by status or server (list_deployments)
One deployment in detail: status, branch, revisions, timings, servers and build settings (get_deployment)

It cannot read or do

Start, retry or roll back a deployment, or change any setting
Server passwords, keys, host names or paths
Environment variables or configuration file contents
Account and billing details
Deployment logs (not readable yet)

Repository addresses are shown without any user name or password.

Lists are capped: up to 200 projects (by name), 200 servers per project, and the 500 most recent deployments of a project. The assistant is told when a list has been cut short, so for an account with more than 200 projects, do not rely on answers that need every project.

Need an assistant that can create deployments or read deployment logs? Use the local MCP server instead. It runs on your own computer with an API key.

You stay in control

The first time your assistant connects, it sends you to DeployHQ to sign in. It acts as you and sees only what you can already see.
It is read-only. It cannot start, retry or roll back a deployment, or change a setting.
An account administrator can switch assistant access off for the whole account under Settings > Account & Access > AI Assistants (MCP). It takes effect immediately. It is on by default, and disabling all AI features for the account also blocks it.
You can disconnect from your assistant at any time, and DeployHQ emails you when an application is approved.

On the approval screen, check that you started the connection and that the access is read-only. DeployHQ recognises Claude; other assistants register themselves, so it cannot verify the name they show.

A plan with API access is required. Accounts that use single sign-on cannot connect assistants that sign in through the browser, such as Claude.ai and ChatGPT, yet. A local assistant such as Claude Code can still use an API key, as the help article explains. Accounts that only allow specific IP addresses block cloud-based assistants such as Claude.ai and ChatGPT. Local assistants connect from your own address.

What you ask an assistant is sent to the company that provides it, so connect assistants you trust. Branch names, project and server names, environments and repository addresses are written by users. An assistant may treat text it reads there as instructions, so take care when an assistant that has other tools acts on what it reads.

Frequently Asked Questions

Can an assistant deploy, retry or roll back through this?

No. The server is read-only. It cannot start, retry or roll back a deployment, and it cannot change any setting. If you want an assistant that can also create deployments or read deployment logs, DeployHQ has a separate MCP server package that runs on your own computer with an API key.

What can the assistant see?

The assistant acts as the DeployHQ user who connected it and sees only what that user can already see. It can read accounts, projects, servers and deployments. It cannot read server passwords or keys, host names or paths, environment variables, configuration file contents, or account and billing details. Repository addresses are shown without any user name or password. Deployment logs are not readable yet.

Who can switch it off?

An account administrator can turn assistant access off for the whole account under Settings > Account & Access > AI Assistants (MCP). It takes effect immediately. It is on by default, and the account-wide option to disable all AI features also blocks it. Any user can disconnect from the assistant at any time, and DeployHQ emails you when an application is approved.

Which assistants can I connect?

Claude (claude.ai), Claude Code, Codex, Gemini CLI, Cursor, VS Code and ChatGPT. In ChatGPT, add it as a custom MCP server from Plugins.

Does it work with single sign-on?

Not yet for assistants that sign in through the browser, such as Claude.ai and ChatGPT. On an account that uses single sign-on they cannot connect at the moment. A local assistant such as Claude Code can still connect with an API key, and the help article explains how.

Do I need a particular plan?

You need a plan with API access. If you do not have an account yet, you can start a free trial.

Read the full documentation No account yet? Try DeployHQ free

Last updated on 29th October 2026