Terminal, Git
Hunt down and fix flaky tests in CI
The coding agent reruns the suite to identify flaky tests, traces each one to a race, shared state or timing assumption, and opens a fix with an explanation.
4 use cases
The coding agent reruns the suite to identify flaky tests, traces each one to a race, shared state or timing assumption, and opens a fix with an explanation.
The agent reads the migration guide, upgrades packages in dependency order, runs tests after each step, and leaves a reviewable commit per package.
Each week the agent takes the most frequent production errors, reproduces them, and opens pull requests with a fix and a regression test.
Feed in the incident channel export and key log excerpts; the agent rebuilds the timeline, separates cause from symptoms, and drafts a blameless postmortem.