Core Concepts

Under Active Development

The content for this page is under active development. It may be incorrect and/or incomplete and is liable to change at any time.


The most important entity in Bubbly's schema is the Release entity. It is the "connection point" for all the data that we want to associate with our Release Readiness decision. Therefore it is very important to know that a release is just an entity, but at the heart of Bubbly.

Check it out in the schema


Getting data in to Bubbly was one of the main challenges we wanted to solve. Adapters were created to solve this.

Adapters are simply Rego files, that define some results from an input. E.g. you can parse a JSON file and convert that into code_issues or a test_run with test_cases.

Check the documentation


Policies are evaluated against each release and form the brains of the Continuous Release Readiness process. Like Adapters, Policies are simply Rego files that define some rules. E.g. you can require that some event occurs in the release process, or deny certain results if they are not good enough.

Check the documentation

GraphQL API#

GraphQL is a query language which is very simple to use, and great for getting structured data out from an API.

For example, to get the head (which is the latest commit on the main branch) for a repository called "demo" we use the following query:

release(where: {has_head_of_with: {name: "demo"}}) {

Check the documentation for more examples


The Bubbly schema has been, and will continue to be, one of the most important parts of the development of Bubbly. As Bubbly is not trying to replace all the rich data sources you might already have, but instead provide a higher-level decision making capability on the data, the schema has to be "just right" in that it doesn't require too much data, but also captures enough to be useful.

One of the main ways in which we solved this simply was to provide a Metadata type, which is basically raw JSON, so that you can add extra data to almost all entities.

Check the documentation which was auto generated from the real schema.