Getting Started

Installation

curl -fsSL https://get.mbm.mn/install | bash

Platform Support

First Run

mbm                    # Launch interactive TUI mbm run "Hello"       # Send a message mbm --help            # Show all commands

License & Activation

License Types

Higher tiers available — view billing page or contact us.

Obtaining a License

1. Sign up at console.mbm.mn/auth
2. Navigate to License & Keys
3. Choose a plan or apply an existing license key

Activate with a License File

mbm license activate /path/to/license.license

Example output:

$ mbm license activate ./license.license
License activated successfully.

  Device ID:  a1b2c3d4e5f6a7b8
  Tier:       pro
  Expires:    2029-12-31

Applying a License Key

License keys use the MBM-XXXX-XXXX-XXXX-XXXX format:

mbm license apply MBM-XXXX-XXXX-XXXX-XXXX

Connecting with an API Key

API keys use the {prefix}-{uuid} format (mf- / mp- / mb- prefix). Apply them inside the TUI:

1. Launch the TUI: mbm
2. Type /connect to open the provider connection dialog
3. Select mbm as the provider
4. Paste your API key when prompted

Alternatively, use the CLI:

mbm providers login # Select "mbm" → paste your API key (mp-abc123...)

Check License Status

mbm license status

  Device ID:    a1b2c3d4e5f6a7b8
  License:      Activated
  Tier:         pro
  Expires:      2029-12-31T00:00:00.000Z
  Max Devices:  1
  Priority Queue: enabled
  API Access:     enabled

Device Binding

Each license is bound to a unique device ID derived from your hardware. If you change machines or hardware, you may need to re-activate your license.

Providers & Models

Supported Providers

Adding API Keys

mbm providers login

Interactive prompt: select provider → paste API key. Credentials are stored in ~/.local/share/mbm/auth.json with restrictive file permissions.

Via Config File

Add credentials to ~/.config/mbm/mbm.json:

{
  "provider": {
    "openai": {
      "apiKey": "sk-proj-..."
    },
    "anthropic": {
      "apiKey": "sk-ant-..."
    }
  }
}

Selecting a Model

mbm models                     # List all available models mbm models openai              # Filter by provider

In the TUI: press / to open the model picker — search by name, filter by provider, select with Enter.

Ollama (Local Models)

Install:

curl -fsSL https://ollama.com/install.sh | sh ollama pull llama3.2

Connect (auto-detect): If Ollama runs on localhost:11434, mbm discovers it automatically.

Custom URL (config):

{
  "provider": {
    "ollama": {
      "baseURL": "https://ollama2.mbm.mn/v1"
    }
  },
  "model": {
    "ollama": "qwen3.6:latest"
  }
}

Environment variable:

export OLLAMA_HOST=http://your-host:11434 mbm

Agent Modes

What Are Agent Modes?

Modes control what the AI agent is allowed to do — read files, run commands, write code, or only analyze. Choose the right mode for your task to stay in control.

SAFE Mode

Requires confirmation for every write operation — file edits, bash commands, git operations. Ideal for learning or cautious exploration.

mbm --safe                # Launch in SAFE mode mbm run --safe "Review the auth module"

Build Mode

Full permissions: file writes, command execution, network access. Best for active coding sessions where you want the agent to implement changes directly.

Plan Mode

Read-only mode: the agent can read files, search code, and fetch documentation — but cannot write, execute, or modify anything. Use before implementation to research and plan.

Permission Configuration

Fine-tune permissions in mbm.json:

{
  "permission": {
    "bash": "ask",
    "edit": "ask",
    "read": "allow",
    "webfetch": "allow",
    "mbm-rag_*": "allow"
  }
}

Values: "allow" — always permit, "deny" — always block, "ask" — confirm each time.

Key Features

RAG — Knowledge Base

Retrieval-Augmented Generation indexes your codebase and documents for semantic search. The AI can answer questions about your entire project.

Setup

mbm rag setup              # Install Docker + pgvector mbm rag enable             # Enable auto-sync on git changes mbm rag index .            # Index current directory

Usage

mbm rag search "how does auth work" mbm rag status             # View KB statistics mbm rag optimize           # Remove duplicates & stale entries

MCP — Model Context Protocol

Connect external tools and data sources via standardized protocol servers.

mbm mcp add               # Add an MCP server
mbm mcp list              # List configured servers
mbm mcp auth <name>       # OAuth authenticate
mbm mcp debug <name>      # Test connection

Sessions

Save, share, fork, and resume conversation sessions.

mbm session list          # List all sessions mbm session share         # Generate share link mbm session compact       # Summarize to reduce context mbm session export        # Export to file

Custom Agents

Create agents with specific permissions, models, and system prompts for specialized tasks.

mbm agent create          # Interactive agent builder mbm agent list            # List your agents

Usage Stats

mbm stats                # Token usage and cost summary

Commands Reference

Top-Level Commands

Global Options

RAG Subcommands

API Keys

Key Format

API keys use the format {prefix}-{uuid}:

Creating Keys

1. Log in to console.mbm.mn
2. Navigate to Account → API Keys
3. Click Create Key
4. Copy the key immediately — it is shown only once

Using Keys in CLI

mbm providers login # Select "mbm" → paste your API key

Or inside the TUI: type /connect, select mbm, and paste your key.

Using Keys in API Requests

curl https://proxy.mbm.mn/v1/chat/completions \
  -H "Authorization: Bearer mf-abc123..." \
  -H "Content-Type: application/json" \
  -d '{"model": "gpt-4o", "messages": [{"role": "user", "content": "Hello"}]}'

Tier Limits

Configuration

Config File

Global config: ~/.config/mbm/mbm.json
Per-project: .mbm/mbm.json (in your project root)

Full Example

{
  "model": {
    "default": "openai/gpt-4o",
    "openai": "gpt-4o",
    "ollama": "llama3.2:latest"
  },
  "provider": {
    "openai": {
      "apiKey": "sk-proj-..."
    },
    "ollama": {
      "baseURL": "http://localhost:11434/v1"
    }
  },
  "permission": {
    "bash": "ask",
    "edit": "ask",
    "read": "allow",
    "webfetch": "allow"
  }
}

Comprehensive Example

Full configuration with custom providers, MCP servers, and advanced options:

{
  "$schema": "https://get.mbm.mn/config.json",
  "provider": {
    "ollama": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Ollama (Local)",
      "options": {
        "baseURL": "http://localhost:11434/v1"
      },
      "models": {
        "qwen3-coder": {
          "name": "qwen3-coder:latest",
          "options": {
            "num_ctx": 262144,
            "stream": false,
            "tools": []
          }
        }
      }
    },
    "openai": {
      "apiKey": "sk-proj-..."
    }
  },
  "mcp": {
    "mbm-rag": {
      "type": "local",
      "command": [
        "python3",
        "~/.mbm/bin/rag/rag_mcp_server.py"
      ]
    }
  },
  "lsp": true,
  "rag_inject": true,
  "permission": {
    "bash": "ask",
    "edit": "ask",
    "read": "allow",
    "webfetch": "allow",
    "mbm-rag_*": "allow"
  }
}

Environment Variables

Directory Layout

Troubleshooting

Ollama Connection Fails

1. Verify Ollama is running: systemctl status ollama or ollama serve
2. Check the port: curl http://localhost:11434/api/tags
3. Pull a model: ollama pull llama3.2
4. If using a remote host, verify the OLLAMA_HOST env var or baseURL in config

License Invalid / Expired

1. Check status: mbm license status
2. Verify your device ID matches the license binding
3. Re-apply the license key: mbm license apply MBM-XXXX-...
4. Visit License & Keys to check your account status

Rate Limit Hit (429 Errors)

1. Check your usage: mbm stats
2. View daily limits at Usage dashboard
3. Consider upgrading your plan for higher limits

RAG Not Working

1. Ensure Docker is running: docker ps | grep mbm-rag
2. Check setup: mbm rag setup
3. Enable auto-sync: mbm rag enable
4. Re-index: mbm rag index .
5. Check status: mbm rag status

API Key Not Working

1. Keys are shown only once at creation — if lost, revoke and create a new one
2. Verify the key format: {prefix}-{32-char-hex}
3. Check if the key was revoked at API Keys
4. Ensure you have an active license for the tier