fleet
An MCP server for Fleet
claude mcp add --transport stdio simplyminimal-fleet-mcp uvx fleet-mcp run \ --env FLEET_READONLY="true" \ --env FLEET_API_TOKEN="your-api-token" \ --env FLEET_SERVER_URL="https://your-fleet-instance.com" \ --env FLEET_ALLOW_SELECT_QUERIES="true"
How to use
Fleet MCP exposes a set of tools that let an AI assistant interact with a Fleet Device Management instance. Through the MCP server, you can query host information, run real-time osquery queries against managed hosts, and manage policies and compliance monitoring. The server also provides capabilities for software inventory, vulnerability tracking, and fleet-level activities/audit logging with optional read-only mode to ensure safety. With the provided configuration, you can connect via your preferred MCP client (for example Claude Desktop, Cursor, Cline, Continue, Zed, Windsurf, etc.) and use the Fleet MCP tools to perform tasks like listing hosts, executing queries, inspecting installed software, and validating policy compliance across your fleet.
How to install
Prerequisites:
- A Fleet instance with API access token permissions
- Python and/or uv installed on your machine (recommended: install uv)
- Access to the internet to pull the fleet-mcp package
Installation steps:
-
Install uv (recommended, as shown in the docs): curl -LsSf https://astral.sh/uv/install.sh | sh
-
Install the fleet-mcp package (via uv) and run the MCP server: uv install fleet-mcp # if using uv to manage packages uvx fleet-mcp run
-
Create or edit your MCP client configuration to point at your Fleet MCP server: { "mcpServers": { "fleet": { "command": "uvx", "args": ["fleet-mcp", "run"], "env": { "FLEET_SERVER_URL": "https://your-fleet-instance.com", "FLEET_API_TOKEN": "your-api-token", "FLEET_READONLY": "true", "FLEET_ALLOW_SELECT_QUERIES": "true" } } } }
-
Run the MCP client configuration with your environment variables in place and ensure the Fleet API token and server URL are correctly set.
Additional notes
Tips and considerations:
- Use the read-only mode (FLEET_READONLY true) for safe experimentation; enable SELECT queries with FLEET_ALLOW_SELECT_QUERIES if you need to run read-only SQL queries.
- Keep API tokens secure; consider using a config file or environment management to avoid embedding tokens in plain text.
- If your client requires a different launcher (e.g., global install), replace the command or adjust the path to fleet-mcp accordingly.
- Ensure network access from the MCP runner to your Fleet instance and that the API token has appropriate permissions for the requested actions.
- The MCP supports multiple clients; reuse the same fleet server configuration across Claude Desktop, Cursor, Cline, Continue, Zed, Windsurf, etc.
Related MCP Servers
nautex
MCP server for guiding Coding Agents via end-to-end requirements to implementation plan pipeline
mcp-yfinance
Real-time stock API with Python, MCP server example, yfinance stock analysis dashboard
pfsense
pfSense MCP Server enables security administrators to manage their pfSense firewalls using natural language through AI assistants like Claude Desktop. Simply ask "Show me blocked IPs" or "Run a PCI compliance check" instead of navigating complex interfaces. Supports REST/XML-RPC/SSH connections, and includes built-in complian
cloudwatch-logs
MCP server from serkanh/cloudwatch-logs-mcp
servicenow-api
ServiceNow MCP Server and API Wrapper
the -company
TheMCPCompany: Creating General-purpose Agents with Task-specific Tools