AI & LLMs Runs locally

Claude Code Hooks Generator

Pick lifecycle events and actions to generate valid Claude Code hooks JSON plus example hook scripts.

ELI5

It writes the settings that run your own checks before and after the AI acts.

A little more detail

What this tool does

Claude Code hooks run shell commands at lifecycle events such as PreToolUse, PostToolUse, and Stop. The JSON shape with matchers and hook arrays is easy to get wrong. This generator builds the settings block and starter scripts for common patterns like blocking dangerous commands or logging tool use.

Use it to

Block risky shell commands before they run
Run a formatter or linter after every file edit
Log agent activity or send notifications when a session stops

Interactive workspace Results update as you type

Hook builder

Event

Matcher

Tool-name regex (e.g. Bash, Edit|Write). Leave empty to match all tools.

Command

Shell command to run. Exit 0 allows or continues; exit 2 blocks on PreToolUse.

Presets

Added hooks

Event	Matcher	Command

Output

Paste the hooks block into .claude/settings.json (project) or ~/.claude/settings.json (user). Hooks run with your full user permissions, so review every command before saving.

Example: block-rm.sh

Blocking PreToolUse hook script. Reads JSON from stdin, checks the Bash command for rm -rf, prints to stderr and exits 2 to block.

Hook definitions and scripts are generated entirely in your browser. Nothing is uploaded or stored on a server.

Claude Code hooks run shell commands at lifecycle events such as PreToolUse, PostToolUse, and Stop. The settings live under the hooks key in settings.json, grouped by event with optional tool matchers and command entries.

PreToolUse and PostToolUse accept a matcher regex to target specific tools. Exit code 2 on PreToolUse blocks the tool call and feeds stderr back to Claude. Use this generator to assemble valid JSON, then save hook scripts locally and point commands at those paths.