SSM AI Integration | Squirrel Servers Manager

🤖

Agent Integration

Allow external AI agents to query system state and execute specific, pre-approved tasks within your SSM instance.

🔒

Secure Playbook Execution

Define exactly which Ansible playbooks agents are permitted to run, ensuring controlled automation.

🔄

Standardized Protocol

Built on the open Model Context Protocol (MCP) for interoperability with various AI agents and platforms.

📊

Resource Access

Provide agents with read-only access to resources like device lists and container information.

Integrating AI with Your Infrastructure Management

Squirrel Servers Manager embraces the future of automation by integrating with the Model Context Protocol (MCP). This allows you to securely connect AI agents (like large language models or specialized automation tools) to your SSM instance, enabling them to interact with your managed infrastructure in a controlled manner.

How it Works

Enable MCP Server: Activate the MCP endpoint in SSM settings.
Generate API Key: Use your API key for your AI agent user.
Configure Agent: Set up your agent to communicate with the SSM MCP endpoint using the API key for authentication.
Define Permissions: Specify exactly which Ansible playbooks the agent is allowed to execute via the SSM settings.
Interact: The agent can now make MCP requests to query information (e.g., list devices, check container status) or invoke allowed tools (e.g., run an approved playbook).

Key Capabilities via MCP

🚀

Execute Allowed Playbooks: Trigger pre-approved Ansible playbooks (executePlaybook).

⏱️

Check Playbook Status: Monitor the progress of running playbooks (getPlaybookStatus).

🖥️

List Devices: Get a list of managed devices (devices://).

🔍

Query Containers: Find container details (findAllContainers, findContainerById).

▶️

Manage Containers: Perform basic actions like start/stop on containers (containerAction).

Security First

MCP integration is powerful, but security is paramount. Always restrict agent permissions by carefully selecting allowed playbooks in the MCP Settings. Never allow all playbooks unless you fully understand the implications. Treat API keys as sensitive credentials.

AI Integration with SSMLeveraging Model Context Protocol (MCP)