feature · testing-debugging-optimization
Taxonomy
Platform extensibility
5 tags · 45 out · 45 in · 3 artifacts · 0 gaps · 1 corpus docs
Testing, debugging & optimization is how you know a change made the agent better: golden datasets, trajectory evals, simulation/replay, online scores, debug UIs, and quality gates in CI/CD — before and after deploy.
Observability tells you what happened; this feature tells you whether a prompt/tool/model change improved outcomes. MIT Index shows most products lack agent-specific eval disclosure.
Part Five owes numbers. Architecture without measurement is not progress. MIT Index: most agents lack agent-specific eval disclosure.
Graph edges for this feature — technologies, category lens, companies, DM.
features/testing-debugging-optimization/README.md
No open gaps for this entity.
Testing, Debugging & Optimization · ok
features/testing-debugging-optimization/README.md
Research path: features/testing-debugging-optimization