GitHub Action that tells you whether a failing test is your fault, a known flaky test or already broken on main. No server, no dependencies.
Open-source Playwright reporter + CLI (on npm) comparing test runs to tell broken tests from flaky ones. Cross-run classification · root-cause rules with suggested checks · failures grouped by shared cause, so you fix once not nine times · in-report triage · spec-file view · trends · CI gate · HTML/JSON/Markdown. Fully offline, no AI.
TestSentry uses AI to analyze failed tests and provide failure categories, explanations, confidence scores, and suggested fixes. Each failure receives a unique fingerprint. The first request calls the AI model; repeated requests use the cached result, reducing duplicate calls, improving speed, and lowering AI usage costs.
Find flaky tests in any language - pytest, Jest, go test, JUnit. Local CLI, no SaaS, no account.
Flaky test detection for GitHub Actions — tells flaky tests from real regressions using JUnit XML, test history, failure fingerprinting and diff blame, with optional LLM triage.
Root-cause diagnosis for flaky tests — identify timing sensitivity, order dependency, parallelism issues, and more.
Self-hosted flaky-test detection for CI: JUnit XML ingestion, statistical flakiness scoring, GitHub issue automation
To associate your repository with the flaky-test-detection topic, visit your repo's landing page and select "manage topics."