Integration Examples

Learn how to integrate nano-claw with various chat platforms and services.

Telegram Integration
Discord Integration
Multiple Provider Setup
Local Model Integration (vLLM)

Telegram Integration

Set up nano-claw as a Telegram bot to chat with your AI assistant on the go.

Prerequisites

A Telegram account
nano-claw installed and configured
Basic nano-claw configuration (API key set up)

Step 1: Create a Telegram Bot

Open Telegram and search for @BotFather
Start a chat and send /newbot
Follow the prompts to choose a name and username for your bot
BotFather will provide you with a token like: 123456789:ABCdefGHIjklMNOpqrsTUVwxyz

Example conversation with BotFather:

You: /newbot
BotFather: Alright, a new bot. How are we going to call it? Please choose a name for your bot.

You: My nano-claw Bot
BotFather: Good. Now let's choose a username for your bot. It must end in `bot`.

You: my_nanoclaw_bot
BotFather: Done! Congratulations on your new bot. You will find it at t.me/my_nanoclaw_bot.
You can now add a description...

Here is your token: 123456789:ABCdefGHIjklMNOpqrsTUVwxyz

Step 2: Configure Telegram in nano-claw

Edit ~/.nano-claw/config.json and add the Telegram configuration:

{
  "providers": {
    "openrouter": {
      "apiKey": "sk-or-v1-YOUR_KEY_HERE"
    }
  },
  "agents": {
    "defaults": {
      "model": "anthropic/claude-opus-4-5",
      "temperature": 0.7,
      "maxTokens": 4096
    }
  },
  "channels": {
    "telegram": {
      "enabled": true,
      "token": "123456789:ABCdefGHIjklMNOpqrsTUVwxyz",
      "allowFrom": ["your_telegram_username"]
    }
  }
}

Configuration Options:

enabled: Set to true to activate the Telegram channel
token: Your bot token from BotFather
allowFrom: (Optional) Array of Telegram usernames that can use the bot. Leave empty for public access.

Step 3: Start the Gateway Server

The gateway server manages all chat channel connections:

nano-claw gateway

Expected Output:

🚀 Starting nano-claw Gateway Server...

✓ Telegram channel initialized
  Bot Username: @my_nanoclaw_bot
  Allowed Users: your_telegram_username

🌐 Gateway server running
Press Ctrl+C to stop

Step 4: Chat with Your Bot

Open Telegram
Search for your bot by username (e.g., @my_nanoclaw_bot)
Start a conversation with /start
Send any message to chat with nano-claw!

Example Conversation:

You: /start
Bot: Hello! I'm your nano-claw assistant. How can I help you today?

You: What's the weather like today?
Bot: I don't have real-time internet access, but I can help you find weather 
information. What's your location?

You: Can you write a Python function for me?
Bot: Of course! What should the function do?

You: Calculate the sum of a list of numbers
Bot: Here's a Python function that calculates the sum of a list:

```python
def sum_list(numbers: list[float]) -> float:
    """Calculate the sum of numbers in a list."""
    return sum(numbers)

# Example usage:
my_numbers = [1, 2, 3, 4, 5]
result = sum_list(my_numbers)
print(result)  # Output: 15
```

Security Tips

Use allowFrom: Restrict bot access to specific users
Keep Token Secret: Never commit your bot token to version control
Use Environment Variables: Store the token in an environment variable:

export TELEGRAM_BOT_TOKEN="123456789:ABCdefGHIjklMNOpqrsTUVwxyz"
nano-claw gateway

Discord Integration

Connect nano-claw to Discord to create a server assistant.

Prerequisites

A Discord account
Administrator access to a Discord server
nano-claw installed and configured

Step 1: Create a Discord Application

Go to Discord Developer Portal
Click "New Application"
Give it a name (e.g., "nano-claw Bot")
Click "Create"

Step 2: Create a Bot User

In your application, go to the "Bot" section
Click "Add Bot"
Confirm by clicking "Yes, do it!"
Under "Token", click "Copy" to copy your bot token
Enable these Privileged Gateway Intents:
- Presence Intent
- Server Members Intent
- Message Content Intent

Step 3: Invite Bot to Your Server

Go to "OAuth2" → "URL Generator"
Select scopes:
- ✅ bot
- ✅ applications.commands
Select bot permissions:
- ✅ Read Messages/View Channels
- ✅ Send Messages
- ✅ Read Message History
Copy the generated URL and open it in your browser
Select your server and authorize the bot

Step 4: Configure Discord in nano-claw

Edit ~/.nano-claw/config.json:

{
  "providers": {
    "openrouter": {
      "apiKey": "sk-or-v1-YOUR_KEY_HERE"
    }
  },
  "agents": {
    "defaults": {
      "model": "anthropic/claude-opus-4-5"
    }
  },
  "channels": {
    "discord": {
      "enabled": true,
      "token": "YOUR_DISCORD_BOT_TOKEN",
      "allowFrom": ["USER_ID_1", "USER_ID_2"]
    }
  }
}

Getting User IDs:

Enable Developer Mode in Discord (User Settings → Advanced → Developer Mode)
Right-click on a user and select "Copy ID"

Step 5: Start the Gateway Server

nano-claw gateway

Expected Output:

🚀 Starting nano-claw Gateway Server...

✓ Discord channel initialized
  Bot: nano-claw Bot#1234
  Servers: 1

🌐 Gateway server running
Press Ctrl+C to stop

Step 6: Use the Bot

In your Discord server:

Mention the bot: @nano-claw help me with something
Direct message: Send a DM to the bot

Example Interaction:

User: @nano-claw What is TypeScript?

nano-claw Bot: TypeScript is a strongly-typed programming language 
that builds on JavaScript. It adds optional static typing, which 
helps catch errors during development and provides better IDE support.

User: @nano-claw Can you create a simple Express server example?

nano-claw Bot: Sure! Here's a simple Express server in TypeScript:
...

Multiple Provider Setup

Use multiple LLM providers for redundancy and flexibility.

Configuration

Configure multiple providers in ~/.nano-claw/config.json:

{
  "providers": {
    "openrouter": {
      "apiKey": "sk-or-v1-YOUR_KEY_HERE"
    },
    "anthropic": {
      "apiKey": "sk-ant-YOUR_KEY_HERE"
    },
    "openai": {
      "apiKey": "sk-YOUR_KEY_HERE"
    },
    "groq": {
      "apiKey": "gsk_YOUR_KEY_HERE"
    },
    "deepseek": {
      "apiKey": "YOUR_KEY_HERE"
    }
  },
  "agents": {
    "defaults": {
      "model": "anthropic/claude-opus-4-5",
      "temperature": 0.7,
      "maxTokens": 4096
    }
  }
}

Usage

nano-claw automatically selects the correct provider based on the model name:

# Uses OpenRouter
nano-claw agent -m "Hello" --model "anthropic/claude-opus-4-5"

# Uses Anthropic directly
nano-claw agent -m "Hello" --model "claude-opus-4-5"

# Uses OpenAI
nano-claw agent -m "Hello" --model "gpt-4-turbo"

# Uses Groq
nano-claw agent -m "Hello" --model "mixtral-8x7b-32768"

Provider Selection Logic

Model Name Pattern	Provider Used
`anthropic/*`	OpenRouter
`openai/*`	OpenRouter
`claude-*`	Anthropic
`gpt-*`	OpenAI
`mixtral-*`	Groq
`deepseek-*`	DeepSeek

Benefits

Redundancy: Fallback if one provider is down
Cost Optimization: Use cheaper models for simple tasks
Feature Access: Access provider-specific models
Testing: Compare different models easily

Local Model Integration (vLLM)

Run nano-claw with local LLM models using vLLM for complete privacy and control.

Prerequisites

Python 3.8+
CUDA-compatible GPU (recommended)
Sufficient VRAM (depends on model size)
Docker (optional, for easier setup)

Option 1: Using Docker

Step 1: Start vLLM Server

docker run --gpus all \
  -p 8000:8000 \
  vllm/vllm-openai:latest \
  --model mistralai/Mistral-7B-Instruct-v0.2

For larger models:

# Llama 3 8B
docker run --gpus all \
  -p 8000:8000 \
  vllm/vllm-openai:latest \
  --model meta-llama/Meta-Llama-3-8B-Instruct

# Mixtral 8x7B (requires ~90GB VRAM)
docker run --gpus all \
  -p 8000:8000 \
  vllm/vllm-openai:latest \
  --model mistralai/Mixtral-8x7B-Instruct-v0.1

Option 2: Manual Installation

Step 1: Install vLLM

pip install vllm

Step 2: Start vLLM Server

python -m vllm.entrypoints.openai.api_server \
  --model mistralai/Mistral-7B-Instruct-v0.2 \
  --port 8000

Configure nano-claw

Edit ~/.nano-claw/config.json:

{
  "providers": {
    "vllm": {
      "apiBase": "http://localhost:8000/v1"
    }
  },
  "agents": {
    "defaults": {
      "model": "mistralai/Mistral-7B-Instruct-v0.2",
      "temperature": 0.7,
      "maxTokens": 2048
    }
  }
}

Usage

# Check if vLLM is running
curl http://localhost:8000/v1/models

# Use with nano-claw
nano-claw agent -m "Hello! Can you introduce yourself?"

Expected Output:

🤖 Agent: Hello! I'm an AI assistant running on your local machine using the 
Mistral-7B model. I can help you with various tasks while keeping all your data 
private and secure. How can I assist you today?

Performance Tips

GPU Selection: Use specific GPUs if you have multiple

CUDA_VISIBLE_DEVICES=0 python -m vllm.entrypoints.openai.api_server ...

Batch Size: Adjust for better throughput

python -m vllm.entrypoints.openai.api_server \
  --model mistralai/Mistral-7B-Instruct-v0.2 \
  --max-num-seqs 256

Quantization: Use quantized models for less VRAM

python -m vllm.entrypoints.openai.api_server \
  --model TheBloke/Mistral-7B-Instruct-v0.2-GPTQ \
  --quantization gptq

Available Local Models

Model	Size	VRAM	Best For
Mistral-7B	7B	~16GB	General use
Llama-3-8B	8B	~18GB	Reasoning
Mixtral-8x7B	47B	~90GB	Complex tasks
CodeLlama-7B	7B	~16GB	Coding
Phi-2	2.7B	~8GB	Fast responses

Benefits of Local Models

✅ Complete Privacy: All data stays on your machine
✅ No API Costs: Free inference after initial setup
✅ No Rate Limits: Use as much as you want
✅ Offline Operation: Works without internet
✅ Customization: Fine-tune models for your needs

Troubleshooting

Out of Memory Error:

Use a smaller model
Enable quantization
Reduce max sequence length
Close other GPU applications

Slow Response:

Check GPU utilization: nvidia-smi
Increase batch size
Use a smaller model
Consider using CPU offloading

Combined Setup Example

Use everything together for maximum flexibility:

{
  "providers": {
    "openrouter": {
      "apiKey": "sk-or-v1-xxx"
    },
    "anthropic": {
      "apiKey": "sk-ant-xxx"
    },
    "vllm": {
      "apiBase": "http://localhost:8000/v1"
    }
  },
  "agents": {
    "defaults": {
      "model": "anthropic/claude-opus-4-5",
      "temperature": 0.7,
      "maxTokens": 4096
    }
  },
  "channels": {
    "telegram": {
      "enabled": true,
      "token": "TELEGRAM_TOKEN"
    },
    "discord": {
      "enabled": true,
      "token": "DISCORD_TOKEN"
    }
  }
}

This setup allows you to:

Chat via Telegram and Discord
Use cloud models for complex tasks
Fall back to local models for privacy-sensitive tasks
Have full redundancy and flexibility

Next Steps

Advanced Features - Custom skills and automation
Use Case Scenarios - Real-world examples
Configuration Guide - Complete reference
Basic Usage - Getting started guide

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Integration Examples

Table of Contents

Telegram Integration

Prerequisites

Step 1: Create a Telegram Bot

Step 2: Configure Telegram in nano-claw

Step 3: Start the Gateway Server

Step 4: Chat with Your Bot

Security Tips

Discord Integration

Prerequisites

Step 1: Create a Discord Application

Step 2: Create a Bot User

Step 3: Invite Bot to Your Server

Step 4: Configure Discord in nano-claw

Step 5: Start the Gateway Server

Step 6: Use the Bot

Multiple Provider Setup

Configuration

Usage

Provider Selection Logic

Benefits

Local Model Integration (vLLM)

Prerequisites

Option 1: Using Docker

Step 1: Start vLLM Server

Option 2: Manual Installation

Step 1: Install vLLM

Step 2: Start vLLM Server

Configure nano-claw

Usage

Performance Tips

Available Local Models

Benefits of Local Models

Troubleshooting

Combined Setup Example

Next Steps

FilesExpand file tree

integration-examples.md

Latest commit

History

integration-examples.md

File metadata and controls

Integration Examples

Table of Contents

Telegram Integration

Prerequisites

Step 1: Create a Telegram Bot

Step 2: Configure Telegram in nano-claw

Step 3: Start the Gateway Server

Step 4: Chat with Your Bot

Security Tips

Discord Integration

Prerequisites

Step 1: Create a Discord Application

Step 2: Create a Bot User

Step 3: Invite Bot to Your Server

Step 4: Configure Discord in nano-claw

Step 5: Start the Gateway Server

Step 6: Use the Bot

Multiple Provider Setup

Configuration

Usage

Provider Selection Logic

Benefits

Local Model Integration (vLLM)

Prerequisites

Option 1: Using Docker

Step 1: Start vLLM Server

Option 2: Manual Installation

Step 1: Install vLLM

Step 2: Start vLLM Server

Configure nano-claw

Usage

Performance Tips

Available Local Models

Benefits of Local Models

Troubleshooting

Combined Setup Example

Next Steps