Telegram AI Bot with n8n

An AI-powered Telegram bot built with the n8n workflow automation platform. It features role-based access control, persistent conversation memory, and integrated DALL-E 3 image generation capabilities.

---

🌟 Features

• Intelligent Conversations: Powered by OpenAI GPT-4 with conversation memory.
• Role-Based Access: Different AI behavior for administrators and regular users.
• Image Generation: Integrated DALL-E 3 for creating images on demand.
• HTML Formatting: Rich message formatting with Telegram-supported HTML tags.
• Conversation Memory: Maintains context across multiple messages using buffer window memory.

📋 Prerequisites

• n8n instance (self-hosted or cloud).
• OpenAI API key (with access to GPT-4 and DALL-E 3).
• Telegram Bot Token (obtain from @BotFather).

🚀 Installation

1. Import Workflow to n8n

1. Download the Telegram_AI_bot_CLEANED.json file from this repository.
2. Open your n8n instance.
3. Go to Workflows → Import from File.
4. Select the downloaded JSON file.
5. Choose "Import as new workflow".

2. Configure Credentials

After import, you'll need to set up the following credentials:

OpenAI API:

1. Go to Credentials in n8n.
2. Create new OpenAI credential.
3. Enter your OpenAI API key.

Telegram Bot:

1. Create a bot via @BotFather.
2. Copy the bot token.
3. In n8n, create new Telegram credential.
4. Paste your bot token.

3. Set Administrator ID

To enable admin privileges:

1. Open the workflow in n8n.
2. Find the Edit Fields node (named "Define Leaders" in some versions).
3. Locate the leader_ids parameter.
4. Replace [0] with your actual Telegram user ID.
   • To find your ID, message @userinfobot.

{
  "name": "leader_ids",
  "type": "json",
  "value": [YOUR_TELEGRAM_ID]
}

4. Activate Workflow

• Click the Active toggle in the top-right corner.
• The bot is now live and ready to receive messages!

---

🏗 Architecture

Workflow Components

• Trigger Node: Listen for incoming events via Telegram webhook listener.
• Authentication Flow:
  • Edit Fields: Defines administrator IDs.
  • Check Leader: Validates user permissions via JavaScript.
  • If Node: Routes requests to the appropriate AI agent.
• AI Agents:
  • AI Agent User: Standard conversational AI for regular users.
  • AI Agent Admin: Priority AI agent for administrators with elevated context.
• Language Model: OpenAI Chat Model (GPT-4-1106-preview) with configurable temperature (0.7) and frequency penalty (0.2).
• Memory: Window Buffer Memory - Maintains the last 10 messages per chat session.
• Tools: Dall-E 3 Tool - Enables image generation via natural language requests.
• Response Handlers:
  • Telegram Node: Sends AI responses back to users.
  • Correct Errors: Fallback handler for HTML formatting issues.

Image Generation Workflow

1. AI agent detects image request intent.
2. Triggers "Execute Workflow Trigger".
3. Generate image in Dall-E 3 (API call to OpenAI).
4. Delivers generated image back to the user via Telegram.

---

💡 Usage

Basic Conversation

Simply message your bot in Telegram. The AI will respond contextually based on your conversation history.

Image Generation

Ask the bot to draw or create images:

• "Draw a sunset over mountains"
• "Create an image of a futuristic city"
• "Generate a picture of a cat wearing a hat"

Admin Commands

Users with admin privileges receive priority handling and specialized system prompts defined in the "AI Agent Admin" node.

---

⚙️ Configuration

Adjusting AI Behavior

Edit these in the OpenAI Chat Model node parameters:

• Temperature (0.0 - 2.0):
  • Lower = more focused and deterministic.
  • Higher = more creative and random.
  • Default: 0.7
• Frequency Penalty (0.0 - 2.0): Reduces repetition in responses. Default: 0.2
• Context Window: Number of previous messages to remember. Default: 10 messages

Custom System Prompts

Modify the AI's personality by editing system messages in:

• AI Agent User node.
• AI Agent Admin node.

---

🛠 Troubleshooting

• Bot doesn't respond: Check if the workflow is Active. Verify Telegram credentials. Check n8n execution logs.
• HTML formatting errors: The Correct errors node handles most issues. Ensure special characters < > & are escaped in custom messages.
• Image generation fails: Verify OpenAI API key has DALL-E 3 access and sufficient credits.

🚀 Customization Ideas

• Add web search capabilities using additional tools.
• Integrate with databases for persistent data.
• Create custom slash commands for specific actions.
• Add file processing capabilities.

---

📜 License

This project is provided as-is for educational and personal use.

🤝 Credits

Built with n8n, OpenAI GPT-4, and Telegram.

---