hillclimb

name	hillclimb
description	Sustained improvement of one named metric toward a target — measure a baseline, then loop hypothesis, implement, re-measure, accept or revert. Use for perf, bundle size, lint count, or test time. A one-off fix is /investigate; this is the loop.
argument-hint	[metric and target, e.g. 'cold start under 400ms']

Hillclimb

One change, one measurement, keep or revert. Never stack untested changes, and never claim a win from reading the diff.

Set up

1. Name one metric and the direction that counts as better. One, not three.
2. Name the target that ends the run.
3. Write the measurement method as a repeatable command and state it. Sample enough to clear the noise — a median of N runs, not one. Freeze it; changing the method invalidates every number taken before the change.
4. Record the baseline, plus a green run of the checks that have to keep passing.

No measurement, no claim.

Loop

Rank the live hypotheses by expected win and take the top one. Each hypothesis names a mechanism ("defer X off the boot path because it blocks first paint"), not "try memoizing something".

• Implement it alone.
• Re-measure with the frozen method and run the checks.
• Accept only when the metric moves past the noise and the checks stay green. Otherwise revert it in full; a tweak that might help does not ride along.
• One commit per accepted win, named with the before and after numbers.

Log every attempt, kept or reverted, so the search accumulates instead of circling.

Stop

At the target, or after three consecutive hypotheses with no accepted win. Don't relax the target to declare victory.

Report

The metric trajectory as a table: attempt, hypothesis, before, after, verdict. The accepted wins. The rejected hypotheses and what each ruled out.