Conduit

Make Your AI Tools Smarter with a Private Knowledge Base

Your Documents. Your AI. Your Control.

Conduit transforms your local documents into a private knowledge base that makes AI tools like Claude Desktop, ChatGPT, Perplexity, and AI coding assistants like Claude Code, Cursor, Copilot, Kiro, and Gemini CLI significantly smarter.

Everything stays local. No documents or artifacts ever leave your machine.

Why Conduit?

AI tools are powerful, but they struggle with:

• Context bloat: Feeding too much information overwhelms the AI
• Missing context: Not enough information leads to hallucinations
• Privacy concerns: Sensitive documents shouldn't leave your machine

Conduit solves this by:

• Intelligent retrieval: RAG (Retrieval-Augmented Generation) and KAG (Knowledge-Augmented Generation) find exactly the right context
• Local-first: All processing happens on your machine - documents never leave
• MCP integration: Works with any AI tool that supports Model Context Protocol

Quick Start (5 minutes)

Install via CLI (Recommended)

curl -fsSL https://raw.githubusercontent.com/amlandas/Conduit-AI-Intelligence-Hub/main/scripts/install.sh | bash

The installer handles everything:

• Installs Conduit CLI and daemon
• Sets up container runtime (Podman/Docker)
• Installs AI models via Ollama
• Configures vector database (Qdrant) and knowledge graph (FalkorDB)
• Auto-configures MCP server in Claude Code

Verify Installation

# Restart terminal, then:
conduit doctor

Add Your Documents

# Add a folder to your knowledge base
conduit kb add ~/Documents/my-project --name "My Project"

# Sync documents (indexes for search)
conduit kb sync

# Test search
conduit kb search "how does authentication work"

That's it! Your AI tools now have access to your private knowledge base.

How It Works

┌─────────────────────────────────────────────────────────────────────┐
│                        Your Local Machine                           │
├─────────────────────────────────────────────────────────────────────┤
│                                                                     │
│   ┌──────────────┐         ┌──────────────────────────────────┐    │
│   │ Your Docs    │ ──────► │        Conduit                   │    │
│   │ (PDF, MD,    │         │  ┌────────────────────────────┐  │    │
│   │  Code, etc.) │         │  │ RAG: Semantic + Keyword    │  │    │
│   └──────────────┘         │  │ KAG: Knowledge Graph       │  │    │
│                            │  └────────────────────────────┘  │    │
│                            └───────────────┬──────────────────┘    │
│                                            │                        │
│                                            │ MCP Protocol           │
│                                            ▼                        │
│   ┌────────────────────────────────────────────────────────────┐   │
│   │              AI Tools (via MCP Server)                     │   │
│   │  Claude Code │ Cursor │ Claude Desktop │ ChatGPT │ etc.   │   │
│   └────────────────────────────────────────────────────────────┘   │
│                                                                     │
└─────────────────────────────────────────────────────────────────────┘

Key Technologies

CLI Reference

Conduit is a CLI-first tool. For the complete command reference, see CLI Command Index.

Essential Commands

# Setup & Health
conduit doctor              # Run diagnostics
conduit status              # Show system status

# Knowledge Base
conduit kb add <path>       # Add document folder
conduit kb sync             # Index documents
conduit kb search <query>   # Search your knowledge base
conduit kb list             # List all sources
conduit kb stats            # Show statistics

# MCP Server
conduit mcp status          # Verify MCP server is configured
conduit mcp configure       # Configure MCP for Claude Code

Search Modes

# Hybrid search (default) - combines keyword + semantic
conduit kb search "authentication flow"

# Semantic search - finds by meaning
conduit kb search "securing user login" --semantic

# Keyword search - exact matches
conduit kb search "OAuth2 client_id" --fts5

Knowledge Graph Queries

# Extract entities from your documents
conduit kb kag-sync

# Query relationships
conduit kb kag-query "Kubernetes"
conduit kb kag-query "authentication" --entities OAuth,JWT --max-hops 2

MCP Server Integration

Conduit automatically creates an MCP server for your knowledge base and configures it for Claude Code.

For Claude Code (Auto-configured)

After running conduit kb sync, your knowledge base is automatically available in Claude Code.

For Other AI Tools (Manual Configuration)

Add to your AI tool's MCP configuration:

{
  "mcpServers": {
    "conduit-kb": {
      "command": "conduit",
      "args": ["mcp", "kb"]
    }
  }
}

Configuration locations:

• Claude Desktop: ~/Library/Application Support/Claude/claude_desktop_config.json
• Cursor: .cursor/mcp.json
• VS Code: .vscode/settings.json (under mcp.servers)

Available MCP Tools

Installation Options

CLI Installation (Recommended)

# Standard install
curl -fsSL https://raw.githubusercontent.com/amlandas/Conduit-AI-Intelligence-Hub/main/scripts/install.sh | bash

# Custom options
curl -fsSL ... | bash -s -- --install-dir ~/.local/bin  # Custom location
curl -fsSL ... | bash -s -- --skip-model                # Skip AI model download
curl -fsSL ... | bash -s -- --no-kag                    # Skip knowledge graph setup
curl -fsSL ... | bash -s -- --verbose                   # Verbose output

Manual Installation

git clone https://github.com/amlandas/Conduit-AI-Intelligence-Hub.git
cd Conduit-AI-Intelligence-Hub
make build
sudo cp bin/conduit bin/conduit-daemon /usr/local/bin/
conduit setup

Desktop App (Experimental)

Note: The Desktop App is currently experimental and under active development. The CLI is the recommended way to use Conduit.

For users who prefer a graphical interface, download the DMG from Releases.

Download

• Apple Silicon (M1/M2/M3/M4): Conduit-x.x.x-arm64.dmg

macOS Security Note

On first launch, you may see "Conduit.app is damaged". Run:

xattr -cr /Applications/Conduit.app

Features (Experimental)

• Dashboard with real-time status
• Knowledge Base management with RAG tuning
• KAG search interface
• Settings and configuration

Supported Document Formats

Uninstalling

The recommended way to uninstall is via the CLI:

# Preview what will be removed
conduit uninstall --dry-run --all

# Uninstall (keeps your data)
conduit uninstall --keep-data

# Full uninstall (removes everything)
conduit uninstall --all

Backup method (if CLI is unavailable):

curl -fsSL https://raw.githubusercontent.com/amlandas/Conduit-AI-Intelligence-Hub/main/scripts/uninstall.sh | bash

Troubleshooting

Quick Diagnostics

conduit doctor    # Comprehensive health check
conduit status    # System status overview

Common Issues

Semantic search shows 0 vectors:

conduit qdrant status              # Check Qdrant
conduit kb sync --rebuild-vectors  # Rebuild vectors

MCP server not working:

conduit mcp status     # Check MCP configuration
conduit mcp configure  # Reconfigure

KAG extraction slow on first run: The extraction model (~4GB) loads on first use. Enable preloading in ~/.conduit/conduit.yaml:

kb:
  kag:
    preload_model: true

For more troubleshooting, see Known Issues.

Documentation

Contributing

We welcome contributions! Please see CONTRIBUTING.md for guidelines.

• Bug Reports: Use the bug report template
• Feature Requests: Use the feature request template
• Questions: Ask in GitHub Discussions

Requirements

The installer handles these automatically:

Privacy & Security

• 100% Local: All documents and processing stay on your machine
• No Telemetry: Conduit doesn't phone home
• Sandboxed Containers: Qdrant and FalkorDB run in isolated containers
• Read-Only MCP: AI tools can only read, not modify your knowledge base

License

MIT License - see LICENSE for details.

Acknowledgments

Built with:

• Qdrant - Vector database
• FalkorDB - Graph database
• Ollama - Local AI models
• Model Context Protocol - AI tool integration