Ansible MCP

Red Hat's MCP server for Ansible Automation Platform exposes the platform to AI agents through six toolsets. Job management tools list available job templates, launch automation jobs and monitor their status while they run. Inventory tools query hosts, check group membership and verify system facts, which is how an agent establishes what a change would actually touch before proposing it. Monitoring tools retrieve job logs, troubleshoot failed tasks and report on the health of the automation environment. User management tools cover access and organizational structure, security and compliance tools let an agent work with managed credentials and verify platform integrity without raw secrets being exposed to the model, and platform configuration tools inspect and tune the environment itself. Deployment is on infrastructure the customer already runs. On RHEL the server is installed through the containerized installer and runs as a pod serving HTTPS on a dedicated port; on OpenShift it is deployed through the Ansible Automation Platform operator. Nothing leaves the environment for a vendor hosted endpoint. Security uses a dual layer model. Server level configuration decides which tools are available at all, including whether the deployment is read only or permitted to launch jobs, and user level permissions apply the role based access control already defined in Automation Platform. An agent therefore cannot launch a template the requesting user could not launch by hand. It fits infrastructure and DevOps teams that want conversational access to automation while keeping enterprise governance in place.