Kapten documentation
Kapten deploys your applications, jobs and databases to Kubernetes clusters in your own cloud account, with environments, pipelines and secrets built in.
Kapten deploys your applications, jobs and databases to Kubernetes clusters that run in your own cloud account. You connect a cluster once, then deploy from a GitHub repository or a container image without writing Kubernetes manifests. Your workloads and data never leave your infrastructure.
Quickstart
Connect a cluster and deploy your first service in a few minutes.
How Kapten works
Organizations, projects, environments, clusters and components.
Guides
Step-by-step instructions for common tasks.
Reference
Resource units, limits and the exact behavior of every setting.
What you can run
| Component | What it is | Typical use |
|---|---|---|
| Application | A long-running service, built from a GitHub repository or pulled from a container image | Web apps, APIs, workers |
| Job | A container that runs to completion, once or on a cron schedule | Migrations, reports, cleanups |
| Database | PostgreSQL, MySQL, MongoDB or Redis, inside your cluster or provisioned by your cloud provider | Application data, caches |
Each environment gets its own copy of these components, its own variables and secrets, and can be deployed in order with a pipeline.
Use these docs with an AI assistant
These docs are built to be read by coding agents as well as people:
/llms.txtlists every page with a one-line summary, and/llms-full.txtholds the whole documentation in one file.- Every page is available as Markdown: append
.mdto its URL, or use Copy as Markdown at the top of the page. - An MCP server at
https://docs.kapten.io/mcplets Claude Code, Cursor, VS Code or Codex search and read the docs directly:
claude mcp add --transport http kapten-docs https://docs.kapten.io/mcp