hookify

What problem does it solve?

Hookify provides lightweight, repository-local guardrails that detect risky CLI commands, unsafe file edits, and problematic prompts, and then warn or block those actions before they run. It reduces accidental destructive operations, leaked secrets, and unsafe changes by surfacing contextual messages and enforcement at the tool level.

Core Features & Use Cases

• Define rules as markdown files with YAML frontmatter to match events and patterns and present human-readable messages when rules trigger.
• Support for single-pattern rules and advanced multi-condition rules that match fields like command, file_path, new_text, and user_prompt using regex and substring operators.
• Use cases: block dangerous rm commands in CI, warn about console.log in TypeScript commits, require checks before stopping an agent, and prevent accidental edits to .env files.

Quick Start

Add a hookify rule that warns on dangerous rm commands by creating a .claude/hookify.dangerous-rm.local.md file with event set to bash, an appropriate regex pattern, and a clear remediation message.