Teaching the lab to learn

A lab that builds labs should also get better at building them. As of today the institute has a self-learning loop in three parts:

• Remember. Corrections and preferences are saved to the agent's memory the moment they happen, so the next session starts out knowing them.
• Improve. A retro skill reviews each substantial session for corrections, friction and steps repeated by hand, and turns them into better rules, skills or hp commands. Small fixes get applied. Bigger ones get proposed.
• Heal. When a documented step breaks, the fix goes into the doc or tool that caused it, not into a one-off workaround.

The loop shouldn't depend on anyone remembering to run it, so a small hook taps the agent on the shoulder once per session after enough work has happened. It's written in shell rather than Rust because it has to keep working while hp itself is mid-surgery.

Plans now go through lavish first, as annotatable pages, before anything gets built.