claude-code-hooks

Automate Claude Code hook setup by defining PreToolUse, PostToolUse, Stop, SessionStart, and SessionEnd events in a structured settings.json file with bash scripts, environment variables, and matchers. This governs validation, formatting, auditing, and lifecycle management across your CI/CD pipeline without manual configuration.

PreToolUse and PostToolUse hooks are event-driven handlers that execute before and after tool execution. PreToolUse applies security guards to block dangerous commands; PostToolUse applies formatters and auditors to validate results. Both use bash scripts and exit codes defined in hook templates to enforce guardrails.

Add a PreToolUse hook with a guard script in settings.json that validates commands against security matchers and exits with a non-zero code to block execution. This governs tool execution and enforces security policy without manual intervention during CI/CD workflows.

Yes. Hooks support validation through PreToolUse matchers, formatting through PostToolUse scripts, and auditing through event handlers. Define bash scripts and environment variables in settings.json to orchestrate security guards, formatters, and auditors across SessionStart, SessionEnd, and UserPromptSubmit events.

You need a structured settings.json file defining hook types, bash scripts, environment variables, matchers, and exit codes, plus hook templates for each event type. These inputs orchestrate security guards, formatters, auditors, and lifecycle management across PreToolUse, PostToolUse, Stop, and session events.

A non-zero exit code blocks execution—PreToolUse hooks reject tool invocation, and PostToolUse hooks mark results as invalid. This enforces validation and security policy in your automation workflow without requiring additional configuration steps.