Can I use this skill to secure shell commands?

Yes, by utilizing the PreToolUse hook, you can route all Bash commands through a validation script or security sandbox before they are executed.

Where does this skill store its configurations?

It manages hook registrations directly within your project's .claude/settings.json file and maintains state in a dedicated .claude/autonomous-state.json file.

How does AC Hooks Manager prevent infinite AI loops?

It implements multiple safety layers, including a 'stop_hook_active' flag to prevent recursive firing, hard iteration limits, and maximum cost thresholds.

What is the purpose of the Stop Hook?

The Stop Hook intercepts Claude Code's termination signal. It allows an analyzer to review the transcript and decide whether Claude should stop or continue with a new instruction, enabling autonomous loops.

AC Hooks Manager

Name: AC Hooks Manager
Author: adaptationio

byadaptationio

Productivity & Workflow

Manages lifecycle hooks for Claude Code to enable autonomous execution, security validation, and automated workflows.

About

AC Hooks Manager provides the foundational infrastructure for transforming Claude Code into a fully autonomous agent. It automates the installation and configuration of specialized hooks, most notably the 'Stop Hook' which intercepts termination signals to allow for task continuation. The skill includes robust safety features such as iteration limits, cost caps, and 'doom loop' prevention. Beyond autonomy, it facilitates PreToolUse security gates and PostToolUse formatting, ensuring that AI-driven development remains safe, consistent, and productive.

Key Features

Comprehensive session management for state and memory persistence
Built-in safety mechanisms for iteration and cost management
Autonomous Stop Hook for continuous multi-step task execution
0 GitHub stars
Automated PostToolUse processing for code linting and formatting
PreToolUse security validation for safe shell command execution

Use Cases

Enforcing security sandboxing for sensitive file and system operations
Building an autonomous coding agent that completes complex features without manual prompts
Standardizing project code quality by triggering formatters after every AI edit

About

Key Features

Comprehensive session management for state and memory persistence
Built-in safety mechanisms for iteration and cost management
Autonomous Stop Hook for continuous multi-step task execution
0 GitHub stars
Automated PostToolUse processing for code linting and formatting
PreToolUse security validation for safe shell command execution

Use Cases

Enforcing security sandboxing for sensitive file and system operations
Building an autonomous coding agent that completes complex features without manual prompts
Standardizing project code quality by triggering formatters after every AI edit