Claude Code Integration¶

Claude Code by Anthropic is a powerful desktop application that brings Claude's advanced AI capabilities to your development workflow. When integrated with Eclair, Claude becomes a data-aware assistant that can discover, analyze, and work with real datasets.

Prerequisites¶

Before setting up Claude Code with Eclair:

✅ Eclair server is installed and running
✅ Node.js is installed on your system
✅ Anthropic account with Claude access

Installation¶

1. Install Claude Code¶

npm install -g @anthropic-ai/claude-code

2. Initial Setup¶

If this is your first time using Claude Code:

claude

Follow the prompts to link your Anthropic account, then exit with Ctrl-C.

Configuration¶

1. Register Eclair MCP Server¶

Add Eclair as an MCP server to Claude Code:

claude mcp add --transport http Eclair http://0.0.0.0:8080/mcp

This command: - Registers Eclair as an MCP server named "Eclair" - Uses HTTP transport to communicate with the server - Connects to your local Eclair instance

2. Copy System Prompt¶

Claude works best with a system prompt that explains Eclair's capabilities:

cp src/eclair/client/claude/claude.md ./CLAUDE.md

This file helps Claude understand how to: - Use Eclair tools effectively - Interpret dataset metadata - Generate appropriate analysis code - Handle data download and processing

3. Verify Configuration¶

Check that Eclair is properly registered:

claude mcp list

You should see Eclair in the list of available MCP servers.

Starting Claude Code¶

1. Ensure Eclair Server is Running¶

# Check server status
curl http://localhost:8080/mcp/health

# Start if not running
eclair-server

2. Start Claude Code¶

claude

You should see the Claude interface with: - Connection to registered MCP servers - Ready prompt for your requests

Claude Code Interface

3. Test Eclair Connection¶

Verify the connection by running:

eclair ping

You should receive a successful ping response from the Eclair server.

Usage Examples¶

Fashion Dataset Analysis¶

Let's walk through the same example as with Gemini CLI:

> Find a fashion-related dataset and visualize some data examples

Step 1: Permission Request Claude will use Eclair to find relevant datasets:

Step 2: Dataset Selection Claude presents dataset options and gets your selection:

Step 3: Data Analysis and Visualization Claude generates comprehensive analysis code: Visualization Code Results

Step 4: Summary Report Claude provides a data scientist-quality summary:

Troubleshooting¶

Common Issues¶

"MCP server not responding"

Check Eclair server status: eclair-server --status
Restart the server: eclair-server
Verify network connectivity

"Permission denied for MCP operations"

Claude asks for permission before using MCP tools
Grant permission when prompted
Check MCP server registration: claude mcp list

"Dataset download failed"

Check internet connectivity
Verify dataset still exists at source
Try alternative datasets
Check for authentication requirements

"Claude Code won't start"

Ensure Node.js is properly installed
Check Claude Code installation: claude --version
Try reinstalling: npm uninstall -g @anthropic-ai/claude-code && npm install -g @anthropic-ai/claude-code

Debug Information¶

Get detailed debug information:

# Check MCP server status
claude mcp status

# View detailed logs
claude --verbose

# Test specific MCP server
claude mcp test Eclair

Configuration Issues¶

Verify your setup:

# List registered MCP servers
claude mcp list

# Check Eclair connection
curl http://0.0.0.0:8080/mcp

# Remove and re-add if needed
claude mcp remove Eclair
claude mcp add --transport http Eclair http://0.0.0.0:8080/mcp