About opum-ai/test-skills
Sustainable test suites for agentic engineering. Six Claude Code skills that stop agent-driven development from burying a project under its own tests, and dig existing projects out: - they set the rules of test economy up front, as a constitution that CI enforces; - they plan the smallest adequate test set for new work; - they measure existing suites with per-test mutation kill matrices; - they cut them down with machine-checked certificates that no fault detection was lost.