Agentic test processes, LLM benchmarks, and other notes on agentic coding (danluu.com)

so nice to read an information-dense somewhat stream-of-thought properly long human written post!

[flagged]