HatMax is a declarative Go service generator. It builds a coherent monorepo of microservices from a single yaml based definition.
HatMax parses the hatmax.yml file to understand the desired services, models, and APIs.
It then scaffolds the directory structure and generates the boilerplate code for each service, including models, repositories, handlers, and configuration.
A command-line interface will help to bootstrap and evolve projects. It can generate the initial yaml template and provides guided commands to extend it in a structured way. You can always edit it manually, but the CLI will ensure consistency as the system grows.
HatMax is built to help you reason clearly and maintain your systems over time, giving you lasting control of your code.
- Declarative: the architecture is defined in a single versioned file with no hidden conventions.
- Standard library first: everything starts with Go’s stdlib and code you own. When a well-established alternative exists, HatMax may offer it as an optional implementation, allowing you to choose the approach that best fits you.
- Transparent code: generated services are plain Go, readable and modifiable, with no hidden layers or reflection magic.
From a single declarative spec, HatMax produces:
- The full monorepo structure.
- Boilerplate for each service: models, repositories, handlers, and application layers
- A presentation layer that can be exposed as a web service.
- Shared operational components: logging, tracing, configuration, metrics, and health checks
- Deployment descriptors: Docker Compose, Nomad, or similar manifests
- Testing scaffolds for unit and integration tests
For detailed information about HatMax's objectives, see Goals.
See more in the Gallery.
It turns out that building a generator also requires building, in parallel, a reference implementation. In this case, it is based on the careful extraction of patterns that have emerged in different personal and commercial projects I have worked on. That process introduces some friction, so I decided to focus on a separate repository, Pulap, which serves as a reference implementation. The intention is not to wait until Pulap is fully complete before bringing updates into HatMax, which would make little sense, but rather to integrate them in batches, each representing a clearly implemented piece of functionality. Several of these updates are pending push and will be published soon.