AnsibleClaw is a framework designed to bridge the gap between Ansible’s extensive library of modules and modern AI agents, such as Cursor, Claude Code, and Gemini CLI. Its primary purpose is to automate the generation of portable skill packages from Ansible module documentation, enabling AI agents to execute infrastructure tasks with concrete, context-aware guidance. By leveraging the output from `ansible-doc`, AnsibleClaw creates comprehensive skill packages that include a SKILL.md file (detailing module parameters and usage), wrapper scripts for CLI and Ansible Automation Platform (AAP) execution, prerequisite checks, and ready-to-use playbooks. This dual-mode approach allows skills to be used both in local CLI environments and in production settings via AAP’s REST API.
At build time, users run the `ansibleclaw` tool to generate skill packages for any Ansible module. The tool can resolve documentation locally or fall back to Ansible Galaxy if collections are not installed. When AAP integration is configured, skill packages can embed controller defaults (such as URL, inventory, credentials, and organization), though sensitive tokens are never included. The generated skills are portable and can be installed directly into agent directories for Cursor, Claude, or Gemini CLI, or distributed as ZIP archives for broader sharing.
The repository includes a web dashboard, accessible via `ansibleclaw ui`, which provides a user-friendly interface for exploring modules, managing collections, generating and composing skills, viewing and deploying skills, and interacting with agents. The dashboard features a collapsible sidebar with sections for module exploration, skill building, skill management, deployment (including AI-powered playbook refinement and AAP integration), and agent interaction (notably with Gemini CLI via a browser-based terminal). This interface streamlines workflows for both novice and advanced users, making skill generation and deployment accessible and efficient.
Each skill package is structured to support both development and production use cases. It contains scripts for running Ansible commands locally, validating prerequisites, publishing playbooks to AAP projects, and interacting with AAP controllers via Python scripts. The assets directory includes example playbooks and requirements files, ensuring that each skill is ready for immediate use. The SKILL.md file serves as the central documentation, guiding AI agents and users through module usage, CLI commands, and AAP workflows.
AnsibleClaw also ships with built-in skills that teach AI agents how to manage Ansible modules, discover available modules, generate new skills on demand, and integrate with AAP. These built-ins are always available, providing foundational capabilities without requiring additional setup. The CLI offers commands for searching modules, generating skills, uninstalling skills from agent directories, and launching the dashboard.
For production environments, AnsibleClaw supports deep integration with AAP, allowing skills to execute ad-hoc commands and job templates through the controller API. Configuration is flexible, with environment variables and YAML files controlling skill output paths, Galaxy API endpoints, collections paths, and AI refinement settings. The project is organized with clear separation between core logic, built-in skills, templates, web components, and documentation, making it easy to extend and maintain.
Overall, AnsibleClaw empowers users and AI agents to leverage Ansible’s automation capabilities in a structured, repeatable, and portable manner, enhancing both development and production workflows across diverse environments.
