Hookify

/hookifyAgents

Every time you correct Claude, you're training it. Hookify captures that training as actual rules. Run it after a frustrating session and watch the same mistakes disappear from every future conversation.

Analyzes your conversation to find behaviors you keep correcting, then writes Claude Code hooks that prevent those behaviors before they happen. The more you use it, the fewer corrections you have to make.

When it runs

User invokes /hookify, asks to prevent a recurring mistake, or wants to create a rule from something Claude did wrong.

Examples

Prompt

/hookify

Outcome

Analyzed conversation, found 2 patterns: Claude added emoji to copy (corrected twice) and used em dashes in content (corrected once). Created hooks to block both behaviors.

Prompt

create a hookify rule to stop Claude from adding TODOs in code

Outcome

Created hook: before any code write, checks for TODO comments and blocks if found with message 'No TODO comments in code, address it now or don\'t.'

agentshooksrulesworkflow

Machine-Readable

This skill is available as structured JSON via the API.

GET /api/v1/skills/hookify

Do-Nothing Score

Find out how close you are to Ghost CEO.

Take the quiz
← Back to Skill LibraryUpdated 2026-03-18