Dawnrose
Log in Start free Menu
/

Docs Getting started

Introduction

Dawnrose is a hosted deployment platform. You point it at a GitHub repository, a Docker image or a folder of files, and it builds the code, runs it and gives it a URL. Databases, volumes, domains, logs, metrics, backups and environments are part of the platform.

Dawnrose is built so that an AI coding agent can do this work for you. Your agent connects over MCP and can create projects, deploy code, wire up databases, read logs and fix failed deploys. Everything it does appears live in the dashboard and is recorded in the activity log under the agent’s name. You can do the same work yourself in the dashboard or with the dawn CLI.

Getting started

Sign up, connect your agent and deploy your first app.

Build and deploy

Repos, folders and images, config as code, variables and templates.

Data and networking

Databases, the private network, domains and HTTPS.

Operate

Environments, logs and alerts, plan limits, and who can do what.

Reference

Every MCP tool and API endpoint, generated from the server.

Help

Answers, what changed, and a person to write to.

How Dawnrose is organized

Everything you run sits in a four-level hierarchy:

workspace → project → environment → service
  • Workspace: the people who share access and their roles (the People page), the plan, billing and usage. Switch workspaces from the top of the sidebar. Every new account gets a personal workspace, and you can be a member of several.
  • Project: a group of services that belong together, such as a web app, its API and its database. A project has a region, picked when you create it: its services run on servers there, or on a server in another region when none there has room.
  • Environment: a separate copy of the project’s services, with its own variables, domains and private network. Each project starts with production. You can add others such as staging, and Dawnrose can create one for each pull request.
  • Service: one thing that runs. That can be an app built from a repository, uploaded files or a single-file function, a Docker image, or a database or app from a template. A service has variables, settings, a deployment history and, if it needs them, a volume, domains and a TCP proxy.

Each time a service is built and rolled out, Dawnrose records a deployment. You can redeploy, roll back to or stop a recent deployment from its history.

Services in the same environment reach each other over a private network at <service>.dawnrose.internal. They share configuration through variable references such as ${{ Postgres.DATABASE_URL }}, and the dashboard draws each reference as a line on the project canvas.

Ways to use Dawnrose

The agent, the dashboard, the CLI and the API all go through the same operations, so a change made in one shows up in the others right away.

Your agent, over MCP

Claude Code, Cursor, Codex, VS Code with Copilot, Windsurf, Gemini CLI, Claude Desktop, opencode and any other client that supports MCP over Streamable HTTP can connect to:

https://app.dawnrose.ai/mcp

The agent authenticates with a token you create in the dashboard. See Connect your agent.

The dashboard

The dashboard at https://app.dawnrose.ai shows each project as a canvas of its services. Open a service to see its deployments, variables, metrics, logs and settings. When an agent is working on a service, the dashboard shows which agent it is and what it is doing.

The dawn CLI

dawn deploys the folder you are in, streams logs, manages variables and domains, and runs local commands with a service’s variables. See dawn CLI.

The API and SDK

Your own code can create projects, deploy code, images and repositories, and read their URLs, through the REST API at https://app.dawnrose.ai/v1 or the TypeScript SDK. A project token keeps that code inside one project. See API and SDK.

Next steps