iceberg MCP Server

Use This MCP server To

Query Apache Iceberg namespaces asynchronously Retrieve tables within specific Iceberg namespaces Fetch detailed schema information for Iceberg tables Access table properties for metadata inspection Integrate Iceberg catalog data into AI workflows Enable LLMs to explore Iceberg data catalogs in real time Log and monitor Iceberg catalog queries for auditing

README

Iceberg MCP

An MCP server for Apache Iceberg catalogs with async and logging.

2025-04-17.10-41-31.mp4

Supported Catalogs

Catalog Type	Supported
Rest Catalogs	✅
AWS Glue	✅
Hive Metastore	❌
S3 Table	❌

Supported Tools

Tools	Description
`namespaces`	Get all namespaces in the Iceberg catalog
`tables`	Get all tables for a given namespace
`table_schema`	Return the schema for a given table
`table_properties`	Return table properties for a given table

Installation

Option 1: Download the Release Binary

Download the latest pre-built binary from the Releases page.

Option 2: Build from Source

To build the project manually, ensure you have Rust installed, then run:

cargo build --release

The compiled binary will be located at: ./target/release/iceberg-mcp

Client Configuration

Claude Desktop

To integrate Iceberg MCP with Claude Desktop:

Open Settings > Developer > Edit Config.
Update claude_desktop_config.json with the appropriate configuration:

Rest Catalogs

{
  "mcpServers": {
    "iceberg-mcp": {
      "command": "PATH-TO-BINARY/iceberg-mcp",
      "env": {
        "CATALOG_KIND": "rest",
        "REST_URI": "http://localhost:8080",
        "LOG_LEVEL": "info"
      }
    }
  }
}

AWS Glue Catalogs

{
  "mcpServers": {
    "iceberg-mcp": {
      "command": "PATH-TO-BINARY/iceberg-mcp",
      "env": {
        "CATALOG_KIND": "glue",
        "AWS_CONFIG_FILE": "/Users/{your_username}/.aws/config",
        "AWS_SHARED_CREDENTIALS_FILE": "/Users/{your_username}/.aws/credentials",
        "PROFILE_NAME": "default",
        "WAREHOUSE": "s3://{bucket_name}/{namespace}/{table}",
        "LOG_LEVEL": "info"
      }
    }
  }
}

Debugging

Claude Desktop

View logs for troubleshooting with:

tail -n 20 -F ~/Library/Logs/Claude/mcp-server-iceberg-mcp.log

iceberg-mcp FAQ

How do I install the iceberg-mcp server?

You can install iceberg-mcp by downloading the release binary or building from source as described in the GitHub repository.

Which Apache Iceberg catalog types does iceberg-mcp support?

It supports Rest Catalogs and AWS Glue, but not Hive Metastore or S3 Table catalogs.

What tools does iceberg-mcp provide for interacting with Iceberg catalogs?

It provides tools to list namespaces, list tables in a namespace, get table schemas, and retrieve table properties.

Does iceberg-mcp support asynchronous operations?

Yes, iceberg-mcp supports asynchronous access to Iceberg catalog data for efficient querying.

Can iceberg-mcp log catalog queries?

Yes, it includes logging capabilities to monitor and audit catalog interactions.

Is iceberg-mcp compatible with multiple LLM providers?

Yes, it is designed to work with various LLMs including OpenAI, Anthropic Claude, and Google Gemini by exposing structured catalog data.

How does iceberg-mcp improve AI workflows?

By providing real-time, structured access to Iceberg metadata, it enables LLMs to reason about and interact with data catalogs effectively.

Can I extend iceberg-mcp to support additional catalog types?

While currently limited to certain catalog types, the server architecture allows for potential extension to other Iceberg catalog implementations.