Habit Hooks, a linter that coaches coding agents

Why it surfaced

Its central idea is unusually clear: a raw metric is a target agents can game, while a linter finding paired with a coaching guide creates an action loop. The site backs that argument with a pre-registered comparison in which coaching produced an 83.3% genuine-fix rate versus 5.6% for a bare complexity metric.

Habit Hooks turns linter violations into short, actionable coaching guides for AI coding agents, with plugins for Python, TypeScript, PHP, and Java.