Writing Hookify Rules

Hookify rules detect dangerous bash commands by matching patterns in YAML frontmatter with name, enabled, event, and pattern fields, plus a Markdown message body that triggers when Claude encounters risky operations like `rm -rf`. Define the pattern to catch the command syntax, set the event to bash, and write a clear warning message.

Yes, Hookify rules can detect sensitive data exposure by defining patterns that match common data leaks—API keys, credentials, tokens—across file edits and prompt submissions. Use pattern matching with optional conditions for multi-condition rules to catch insecure code patterns before they're committed.

Hookify rules enforce coding standards by automatically detecting violations like console.log in production TypeScript, unused imports, or non-compliant patterns across bash commands and file edits. Configure rules with YAML frontmatter and trigger messages so Claude warns your team without manual review overhead.

Hookify rules support both simple single-pattern rules and advanced multi-condition rules through optional conditions fields in YAML frontmatter. Multi-condition rules let you detect nuanced risks—for example, flagging risky commands only in specific file contexts—beyond basic pattern matching.

Set the event field to file in your YAML frontmatter and define a pattern matching the code you want to catch—like production logging statements. When Claude edits a file, the pattern triggers and displays your Markdown message, stopping unsafe changes before they're written.

Hookify rules can apply to stop events to block risky workflow steps and completion checks. Configure the event field to stop, set your safety pattern, and provide a Markdown message explaining why the step was blocked, preventing unsafe operations at any stage in your process.