Can We Classify Flaky

Analyze test suites for flaky tests using LLM-based classification with context-augmented reasoning. Applies findings from Berndt et al. (2026) showing that test code alone is insufficient — the skill teaches Claude to gather surrounding project context (configs, dependencies, environment, production code) before classifying. Trigger phrases: 'find flaky tests', 'classify flaky tests', 'detect test flakiness', 'why is this test flaky', 'analyze test reliability', 'flaky test triage'

ndpvt-web Updated 14 repo stars

File contents

ndpvt-web/arxiv-claude-skills commit 1a2fa6d235

Frequently asked questions

npx get-skill add ndpvt-web/can-we-classify-flaky