Getting Started¶

Welcome to haive-mcp! Get up and running with 1900+ MCP servers in minutes.

Quick Install¶

Install haive-mcp with a single command:

pip install haive-mcp

That’s it! No complex setup, no configuration files. The system automatically discovers and installs MCP servers as needed.

Your First Agent¶

Create an MCP-enabled agent in just a few lines:

import asyncio
from haive.mcp.agents.enhanced_mcp_agent import EnhancedMCPAgent
from haive.core.engine.aug_llm import AugLLMConfig

async def main():
    # Create agent with automatic MCP discovery
    agent = EnhancedMCPAgent(
        name="my_assistant",
        engine=AugLLMConfig(temperature=0.7),
        mcp_categories=["core"],  # Auto-install filesystem, database, search
        auto_install=True
    )

    # Initialize (discovers and installs needed servers)
    await agent.initialize_mcp()

    # Use the agent - tools are automatically available!
    result = await agent.arun("List all Python files in the current directory")
    print(result)

asyncio.run(main())

Key Features¶

Zero Configuration¶

Haive-mcp works out of the box:

No manual installation of MCP servers
No configuration files to write
No tool registration needed
Automatic discovery from 1900+ servers

Dynamic Discovery¶

The agent analyzes your task and automatically finds the right tools:

# Agent automatically discovers file tools for this task
await agent.arun("Organize my documents by date")

# Agent automatically discovers web search tools for this task
await agent.arun("Research the latest AI developments")

# Agent automatically discovers database tools for this task
await agent.arun("Query the customer database for recent orders")

Rich Ecosystem¶

Access to 1900+ MCP servers across categories:

Core Tools: Filesystem, database, web search
AI Enhanced: LLM integrations, embeddings, analysis
Developer Tools: GitHub, code analysis, documentation
Data Tools: Processing, visualization, transformation
Communication: Email, Slack, notifications
And much more!

Common Use Cases¶

File Management¶

agent = EnhancedMCPAgent(
    name="file_manager",
    engine=AugLLMConfig(),
    mcp_categories=["enhanced_filesystem"],
    auto_install=True
)
await agent.initialize_mcp()

# The agent handles complex file operations
await agent.arun("Organize all images by date taken and remove duplicates")

Data Analysis¶

agent = EnhancedMCPAgent(
    name="analyst",
    engine=AugLLMConfig(),
    mcp_categories=["core", "ai_enhanced"],
    auto_install=True
)
await agent.initialize_mcp()

# Analyze data with automatic tool selection
await agent.arun("Analyze sales.csv and create a trend report with visualizations")

Web Research¶

agent = EnhancedMCPAgent(
    name="researcher",
    engine=AugLLMConfig(),
    mcp_categories=["core", "browser_automation"],
    auto_install=True
)
await agent.initialize_mcp()

# Research with web tools
await agent.arun("Research competitors' pricing strategies and create a comparison")

Understanding Categories¶

MCP servers are organized into logical categories:

core: Essential everyday tools - filesystem, database, web search
ai_enhanced: AI-powered capabilities - LLM chains, embeddings, analysis
enhanced_filesystem: Advanced file operations - bulk processing, smart organization
time_utilities: Date and scheduling - calendars, reminders, time zones
browser_automation: Web interaction - scraping, testing, automation
crypto_finance: Financial tools - trading, blockchain, analytics
github_enhanced: Development tools - code analysis, PR management
notifications: Communication - alerts, emails, messaging

Prerequisites¶

System Requirements¶

Python 3.8 or higher
Node.js 18+ (for NPM-based servers)
2GB RAM minimum
Internet connection (for auto-discovery)

Optional Components¶

These enhance functionality but aren’t required:

Docker (for containerized servers)
Git (for repository-based servers)
Redis (for distributed caching)

Verification¶

Verify your installation:

from haive.mcp import verify_installation

# Check installation
status = verify_installation()
print(f"Installation status: {status}")

# List available categories
from haive.mcp.discovery import list_categories
categories = list_categories()
print(f"Available categories: {categories}")

# Check server count
from haive.mcp.registry import count_available_servers
count = count_available_servers()
print(f"Available servers: {count}")  # Should show 1900+

Troubleshooting¶

Common Issues¶

Import Error: Ensure haive-mcp is installed: pip install haive-mcp
Node.js Not Found: Install Node.js from https://nodejs.org/ (v18+ required)
Slow First Run: Initial discovery and caching takes time. Subsequent runs are faster.
Connection Timeout: Check internet connection. Increase timeout: initialization_timeout=120

Debug Mode¶

Enable debug output for troubleshooting:

import logging
logging.getLogger("haive.mcp").setLevel(logging.DEBUG)

# Create agent with verbose output
agent = EnhancedMCPAgent(
    name="debug",
    engine=AugLLMConfig(),
    mcp_categories=["core"],
    auto_install=True,
    verbose=True
)

Getting Help¶

Resources¶

Documentation: Full docs at Haive MCP Documentation
Tutorials: Step-by-step guides at Tutorials
Examples: Code samples at Examples
API Reference: Detailed API at autoapi/haive/mcp/index

Community¶

GitHub Issues: Report bugs and request features
Discussions: Ask questions and share experiences
Discord: Real-time help from the community

Next Steps¶

Now that you’re up and running:

Follow the tutorials - Tutorials for hands-on learning
Explore examples - Examples for real-world patterns
Read the guides - Guides for best practices
Check the API - autoapi/haive/mcp/index for detailed reference

Welcome to the world of 1900+ MCP servers at your fingertips!