Llm Gate

This adds AI-powered quality gates to your workflow using Claude Code's prompt hooks. Instead of writing complex regex or scripts, you write a natural language prompt that validates operations like git commits, file writes, or destructive commands. A small LLM (Haiku by default) checks each operation and blocks it if something's wrong. The conventional commit validator and API key leak prevention examples are solid starting points. It's clever for catching things static tools miss, like whether a commit message actually makes sense or if you're about to nuke an important directory. Just don't go overboard with gates or you'll slow yourself down waiting for validation on every action.

skills.sh installs: 158
GitHub stars: 2.9K
last updated: Jun 8, 2026

View on GitHub

Install

$ npx -y skills add rohitg00/pro-workflow --skill llm-gate --agent claude-code

Installs into .claude/skills of the current project.

Files

LLM Gate

Use Claude Code's type: "prompt" hooks to create intelligent quality gates that use AI to verify operations.

Trigger

Use when:

Setting up commit message validation
Enforcing code conventions beyond what linters catch
Creating smart guardrails for specific operations

How Prompt Hooks Work

Claude Code supports hooks with type: "prompt" that run a small LLM (Haiku by default) to verify conditions:

{
  "PreToolUse": [{
    "matcher": "Bash",
    "hooks": [{
      "type": "prompt",
      "if": "Bash(git commit*)",
      "prompt": "Check if this git commit follows conventional commit format (<type>(<scope>): <summary>). The commit command is: $ARGUMENTS. Return {\"ok\": true} if valid, {\"ok\": false, \"reason\": \"...\"} if not.",
      "model": "haiku",
      "timeout": 15
    }]
  }]
}

The hook: