>_ AI Agents

Your AI Agent's
Deployment Layer

DeployHQ gives AI coding agents a complete deployment platform. Your agent pushes code — DeployHQ handles builds, transfers, and rollbacks. Works with Claude Code, Codex, Cursor, GitHub Copilot, and more.

curl -sSL deployhq.com/install/cli | sh
brew install deployhq/tap/dhq
Works with any agent
CLI + MCP access
JSON output for agents

Last updated on 4th April 2026

Terminal
$ curl -sSL deployhq.com/install/cli | sh
✓ dhq installed successfully
 
$ dhq hello
 
Welcome to DeployHQ CLI! 👋
Logged in as you@company.com
3 projects • 5 servers • 142 deployments
Ready to deploy.

AI coding agents can write code — but they need infrastructure to deploy it. DeployHQ is the deployment platform that bridges the gap. Agents connect via the CLI or MCP server, trigger deployments with structured JSON output, and manage the full lifecycle from project creation to rollback. No custom integrations needed.

Works With Your Favourite Agents

Bring your agent — DeployHQ is the deployment platform underneath.

How Agents Connect

The DeployHQ CLI is the primary interface — a single binary with zero dependencies.

JSON Output

Every command supports --json with optional field selection. Get machine-readable responses perfect for agents and automation scripts.

Zero Dependencies

A single binary with no runtime dependencies. Download it, run it. Works on macOS, Linux, and Windows out of the box.

30+ Commands

Manage projects, servers, deployments, rollbacks, environment variables, SSH keys, and more. Full coverage of the DeployHQ API.

AI Assistant

Built-in local AI assistant powered by Ollama. Ask questions about your deployments in natural language — all data stays on your machine.

CI/CD Integration

Authenticate via environment variables for seamless GitHub Actions integration. No interactive login needed in pipelines.

MCP Server

Agents that support MCP can discover DeployHQ tools automatically. The CLI is the recommended interface — MCP provides an alternative integration path.

Up and Running in Minutes

Three steps to deploy from the command line.

1

Install

Download the binary or install via Homebrew.

$ brew install deployhq/tap/dhq
✓ dhq installed successfully
2

Authenticate

Log in with your DeployHQ credentials.

$ dhq auth login
✓ Authenticated as you@company.com
3

Deploy

Start deploying from your terminal.

$ dhq deploy -p my-app --wait
✓ Deployed to production in 42s

Built for AI Agents

First-class support for autonomous deployment workflows.

Machine-Readable Output

Every command supports --json with optional field selection. Get exactly the data your agent needs.

Breadcrumb Navigation

Agents discover available commands and flags at every step. No guessing, no hallucinating endpoints.

144+ API Endpoints

Full coverage of the deployment lifecycle — projects, servers, deploys, rollbacks, environment variables, and more.

agent session
$ dhq project list --json --fields name,permalink
[{"name":"magic-bean","permalink":"magic-bean"}]
$ dhq deploy create -p magic-bean -s production --wait
Preparing deployment...
Building (npm run build)...
Transferring 42 files...
✓ Deployed to production in 38s
$ dhq deploy show -p magic-bean --json --fields status,duration
{"status":"completed","duration":"38s"}
Read the full documentation →

Frequently Asked Questions

Which AI agents work with DeployHQ?

Any agent that can run shell commands can use the DeployHQ CLI. This includes Claude Code, OpenAI Codex, Cursor, GitHub Copilot, Windsurf, Cline, and others. Agents that support MCP can also discover DeployHQ tools automatically.

How do AI agents connect to DeployHQ?

Agents connect primarily through the DeployHQ CLI (dhq). Every command supports --json output with optional field selection, making responses machine-readable. Agents authenticate via environment variables and can access 144+ API endpoints for full deployment lifecycle management.

What is the DeployHQ MCP server?

The MCP (Model Context Protocol) server allows compatible agents like Claude Code to discover and use DeployHQ tools automatically. The CLI is the recommended primary interface — MCP provides an alternative integration path for agents that support it.

Can I use DeployHQ without an AI agent?

Absolutely. DeployHQ works through the web dashboard, the CLI, or AI agents. The CLI also includes a built-in AI assistant powered by Ollama that runs entirely on your machine for local help with commands and deployments.

Does it work in CI/CD pipelines?

Yes. The CLI supports authentication via environment variables (DHQ_EMAIL and DHQ_API_KEY), so you can use it in GitHub Actions and other CI/CD systems without interactive login.

Is it free?

The CLI itself is free and open source. You need a DeployHQ account to use it — free plans are available with paid plans starting at affordable rates.

Ready to let your agent deploy?

Get started with DeployHQ and connect your AI agent in minutes.

10-day free trial • No setup fees • Cancel anytime

Get started today for just $9/month

That's unlimited deployments and 3 projects.

Start your free 10 day trial