Illumio MCP Server

What is Illumio MCP Server?

A Model Context Protocol (MCP) server that provides an interface to interact with Illumio PCE (Policy Compute Engine). It enables programmatic access to Illumio workload management, label operations, and traffic flow analysis via conversational AI.

How to use Illumio MCP Server?

Requires Python 3.8+, access to an Illumio PCE instance, and valid API credentials. Configure via environment variables (PCE_HOST, PCE_PORT, PCE_ORG_ID, API_KEY, API_SECRET). Run using uv and integrate with Claude Desktop by adding the server configuration to claude_desktop_config.json.

Key features of Illumio MCP Server

• Workload management: create, update, delete workloads
• Label operations: create, delete, get labels
• Traffic analysis: get flows and summaries with filters
• Policy management: get rulesets with optional filtering
• IP lists management: get IP lists by name, description, or ranges
• Connection testing and event retrieval with filtering

Use cases of Illumio MCP Server

• Manage workloads and labels in an Illumio PCE
• Retrieve and analyze traffic flow summaries for security analysis
• Check PCE connectivity and health status
• View and filter events by type, severity, or status
• Query and filter rulesets and IP lists for policy audits

FAQ from Illumio MCP Server

What runtime dependencies are required?

Python 3.8 or later, network access to an Illumio PCE instance, and valid API key and secret credentials.

How do I configure the server for Claude Desktop?

Add the server entry under mcpServers in claude_desktop_config.json, specifying the uv command, the server directory, and environment variables for PCE connection details.

Are resources fully implemented?

No. Resources are noted as unfinished and are planned for a later update.

What transport/authentication does the server use?

The server uses an API key and secret passed as environment variables to authenticate with the Illumio PCE over HTTPS.

Where does the data live?

All data resides in the user’s Illumio PCE instance; the server only reads and writes through its API and does not store data locally.