CI is the new bottleneck. Everyone's answer is a faster runner. The cheaper answer: stop running tests your diff can't reach. Change one file in hono and 7 of its 137 test files can see it. The other 130 run anyway, on every PR, on your bill. So I built shortlist. It reads the repo's import graph, walks it backwards from your diff, and runs only the test files that can reach the change. Vitest, Jest, node:test, Mocha, Bun, pytest. Monorepos get a runner per package. When it can't be sure (a lockfile, a dependency bump, the test workflow itself), it runs everything and tells you why. Never a guess. One line in GitHub Actions: uses: codegobrrrr9/shortlist@v1 Or paste into your agent: Install the shortlist skill from github.com/codegobrrrr9/shor…, refer to the repo's AGENTS.md.

Oct 6, 2026 · 2:15 AM UTC

14
11
51
7,075
Sort replies: Relevant Recent Liked
Replying to @neil_xbt
Huge win for monorepos
1
20
Replying to @neil_xbt
7 out of 137 is a brutal number for whoever's paying the CI bill. What does it do when the diff touches a shared fixture or config?
13
Replying to @neil_xbt
I'd have bet an import graph would miss enough to make skipping tests risky. When a dependency bump forces a full run, does shortlist name the package that caused it, or just say it fell back?
7
Replying to @neil_xbt
Good idea. Running everything when it is not sure, and telling you why, builds trust. How much CI time did it save on a real repo so far?
16
Replying to @neil_xbt
Test selection is the right lever. The risk is the edges an import graph can't see: fixtures, env vars, snapshot files, config read at runtime. Keeping a full run on main or nightly means a missed edge shows up within hours instead of after a release.
21
Replying to @neil_xbt
Running only tests the import graph can reach is such an obvious win. How do you handle dynamic imports?
9
Replying to @neil_xbt
i would check how shortlist treats config files, fixtures and dynamic imports. a changed file with no static path to a test can still break it, so a small always-run set as a safety net might be worth keeping.
19
Replying to @neil_xbt
neil, 7 of 137 hono test files actually seeing a one-file change is a brutal stat, shortlist makes a lot of sense. i'm chloe, building Audiee. wanna connect.
13