Posts tagged #testing

Evaluation-driven development for agents: a regress-gate that can't fail your build

August 15, 20269 min

#ai-agents
#testing

Agents need tests the way code does, but the assertion is fuzzy. Here's the EDD loop I built into ReplayGate: deterministic offline replay that gates the build with real exit codes, an LLM judge that only ever advises, and the honest cost of pinning replay to a hash.

Finding the N+1 before it finds prod: the benchmark that couldn't fail

August 13, 20269 min

#django
#orm

I wrote a benchmark to prove my vault API's eager loading worked, charted it, and published it. Five weeks later I deleted the eager loading from the viewset to watch the benchmark go red. Fifteen tests passed in 4.67 seconds and the benchmark cheerfully printed the same flat line.