Fleetp MCP

What is Fleet MCP?

Fleet MCP is a Model Context Protocol server that enables AI assistants to interact with Fleet Device Management for device management, security monitoring, and compliance enforcement. It provides tools for host management, live osquery queries, policy management, and software inventory tracking.

How to use Fleet MCP?

Install via uvx fleet-mcp run or pip install fleet-mcp, then configure your MCP client with Fleet server URL and API token. Set environment variables like FLEET_READONLY and FLEET_ALLOW_SELECT_QUERIES to control access. See client-specific configuration examples for Claude Desktop, Cursor, Cline, Continue, Zed, Windsurf, VS Code, Sourcegraph Cody, Augment Code, and LM Studio.

Key features of Fleet MCP

• Host management: list, search, query, and manage hosts across your fleet
• Live osquery execution: run real-time queries against hosts
• Policy management: create, update, and monitor compliance policies
• Software inventory: track installations and vulnerabilities across devices
• Team and user management: organize hosts and users into teams
• Read-only mode with optional SELECT-only query execution

Use cases of Fleet MCP

• IT teams performing device inventory and compliance audits across a managed fleet
• Security analysts running live osquery queries to detect anomalies or threats
• Developers automating device provisioning, software tracking, and policy enforcement
• Administrators monitoring fleet activities and reviewing audit logs

FAQ from Fleet MCP

What is read-only mode?

Read-only mode (FLEET_READONLY=true, the default) restricts all tools to read-only operations, preventing any modifications to Fleet state. You can also allow SELECT queries separately with FLEET_ALLOW_SELECT_QUERIES=true.

What dependencies does Fleet MCP require?

Fleet MCP requires a running Fleet Device Management instance and a valid Fleet API token. It can be installed via pip or uv, with uv being recommended for development.

Where does Fleet MCP store data?

Fleet MCP does not store data locally; it communicates with your Fleet server via its API. Configuration can be stored in a TOML file, but sensitive tokens should be kept secure with restricted file permissions (e.g., chmod 600).

Which MCP clients are supported?

Fleet MCP works with Claude Desktop, Cursor, Cline, Continue, Zed, Windsurf, VS Code, Sourcegraph Cody, Augment Code, LM Studio, and any generic MCP-compatible client using the stdio transport.

How do I secure my Fleet API token?

Use a configuration file with --config flag instead of embedding tokens in environment variables, set file permissions to 600, enable read-only mode by default, and regularly rotate API tokens.