How to Set Up Lifecycle Hooks for Automated Enforcement in Claude Code

Mar 21, 2026

Problem

When I wrote rules in CLAUDE.md like “always run typecheck before committing,” Claude sometimes forgot to follow them. I kept adding more rules, but compliance didn’t improve.

The core issue: rules are text that requests behavior. Claude processes them as suggestions, not guarantees.

Rule in CLAUDE.md: "Always run typecheck before committing"
Claude's behavior: Sometimes remembers, sometimes forgets
Result: Inconsistent code quality

Environment

Claude Code (latest)
TypeScript project
CLAUDE.md with 150+ lines of rules

What happened?

I noticed that rules at line 137 of a 200-line CLAUDE.md were often ignored. When I investigated, I found:

- LLMs process rules as suggestions, not guarantees
- Long CLAUDE.md files push older rules out of context
- The model may "decide" the rule doesn't apply

One Reddit user put it perfectly: “One of the statistically possible results of saying ‘don’t commit without linting first’ is for the tool to decide to commit without linting first.”

How to solve it?

I replaced rules with lifecycle hooks.

The Hook Types

Claude Code supports three hook types:

Hook Type	When It Runs	Use Cases
PreToolUse	Before tool execution	Validation, blocking dangerous actions
PostToolUse	After tool execution	Auto-formatting, type checking, linting
Stop	When session ends	Final verification, audit logging

Solution #1: PostToolUse Typechecking Hook

I added a hook that runs TypeScript type check after every edit:

{
  "hooks": {
    "PostToolUse": [
      {
        "matcher": "Edit|Write",
        "hooks": [
          {
            "type": "command",
            "command": "npx tsc --noEmit"
          }
        ]
      }
    ]
  }
}

Now Claude doesn’t choose to be typechecked. The environment enforces it.

Solution #2: PreToolUse Git Push Review

I added a hook that opens a diff review before git push:

{
  "hooks": {
    "PreToolUse": [
      {
        "matcher": "Bash",
        "hooks": [
          {
            "type": "command",
            "command": "echo 'Review changes before push' && code --wait --diff"
          }
        ]
      }
    ]
  }
}

Solution #3: Comprehensive Hook Configuration

Here’s my full hook setup:

{
  "hooks": {
    "PreToolUse": [
      {
        "matcher": "Bash",
        "hooks": [
          {
            "type": "command",
            "command": "tmux-reminder.sh"
          }
        ]
      }
    ],
    "PostToolUse": [
      {
        "matcher": "Edit|Write",
        "hooks": [
          {
            "type": "command",
            "command": "prettier --write $FILE_PATH"
          }
        ]
      },
      {
        "matcher": "Edit|Write",
        "glob": "*.ts",
        "hooks": [
          {
            "type": "command",
            "command": "npx tsc --noEmit"
          }
        ]
      }
    ],
    "Stop": [
      {
        "type": "command",
        "command": "audit-console-logs.sh"
      }
    ]
  }
}

The reason

I think the key reason hooks work better than rules is:

Deterministic enforcement: Hooks run every time, regardless of context
No memory required: The hook doesn’t need Claude to remember it exists
Infrastructure vs documentation: Rules are text hoping for compliance; hooks are code guaranteeing compliance

After implementing hooks, my review time dropped dramatically. By the time I looked at the code, the structural problems were already gone.

Summary

In this post, I showed how to set up lifecycle hooks in Claude Code for automated enforcement. The key point is: rules degrade, hooks don’t. Instead of hoping Claude remembers your rules, build infrastructure that enforces them automatically.

Final Words + More Resources

My intention with this article was to help others share my knowledge and experience. If you want to contact me, you can contact by email: Email me

Here are also the most important links from this article along with some further resources that will help you in this scope:

👨‍💻 Reddit: Claude AI Discussion

Oh, and if you found these resources useful, don’t forget to support me by starring the repo on GitHub!