HeytalePazguato

This matches what I've seen working with automated systems. The watching part is genuinely underrated. Evals give you a score. Watching gives you intuition about failure modes you didn't know to test for.

Sitting with a running system teaches you things you would never think to measure.

The camps framing misses something. After 15 years writing code I've found the people who ship the best products understand both sides. You need to care about craft enough to know when AI output is garbage, and you need to care about shipping enough to not gold-plate things that don't matter.

The slop problem isn't AI, it's people who can't tell the difference between good and bad output because they never developed the craft in the first place. AI just makes that gap more visible.

The hooks system is the most underappreciated thing in what leaked. PreToolUse, PostToolUse, session lifecycle, all firing via curl to a local server. Clean enough to build real tooling on top of without fighting it.

The frustration regex is funny but honestly the right call. Running an LLM call just to detect "wtf" would be ridiculous.

KAIROS is what actually caught my attention. An always-on background agent that acts without prompting is a completely different thing from what Claude Code is today. The 15 second blocking budget tells me they actually thought through what it feels like to have something running in the background while you work, which is usually the part nobody gets right.