Hosted predictive test selection products solve a real problem, and they come with a procurement conversation: another vendor, another bill, and a security review about sending build and test metadata to someone else's servers. Some teams clear that review. Plenty do not, or cannot justify the budget for a suite that is only slow, not on fire.

There is a smaller, older idea that needs none of it: test impact analysis from a saved dependency map, stored wherever you already store build state. This post shows a complete self-hosted CI architecture for it with blastradius , using either the GitHub Actions cache or a private S3 bucket. I ran both paths end to end on fresh clones so the setup, and the places it breaks, are tested rather than described.

What actually gets stored

Before you decide where it can live, look at what it is. After a base-branch build, blastradius writes an index that maps each test to the production classes it loaded.…