https://github.com/anlaki-py/akibot

AkiBot - The Ultimate Telegram AI Assistant
https://github.com/anlaki-py/akibot

ai assistant-chat-bots gemini telegram telegram-bot

Last synced: 7 months ago
JSON representation

AkiBot - The Ultimate Telegram AI Assistant

• Host: GitHub
• URL: https://github.com/anlaki-py/akibot
• Owner: anlaki-py
• License: mit
• Created: 2025-01-12T16:08:24.000Z (9 months ago)
• Default Branch: main
• Last Pushed: 2025-02-17T11:40:17.000Z (8 months ago)
• Last Synced: 2025-02-17T12:32:32.791Z (8 months ago)
• Topics: ai, assistant-chat-bots, gemini, telegram, telegram-bot
• Language: Python
• Homepage: https://anlaki-py.github.io/AkiBot/
• Size: 245 KB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# AkiBot (v1.0.8) - The Ultimate Telegram AI Assistant

This is a Telegram bot powered by the Gemini AI model, designed to interact with users through various media types and provide helpful responses.

## Functionality

The bot is capable of:

-   **Text Interaction:** Engaging in conversational text-based chats, maintaining context throughout the conversation.

-   **Image Understanding:** Processing images sent by users, allowing for vision-based queries.

-   **Document Analysis:** Reading and understanding text-based documents.

-   **Audio Processing:** Handling audio messages and files.

-   **Instagram Media Downloading:** Downloading photos and videos from Instagram links.

-   **YouTube to MP3 Conversion:** Converting YouTube videos to MP3 audio files.

-   **Webpage to Markdown Conversion:** Converting webpages to Markdown format.

-   **YouTube Transcript Extraction**: Extracting transcripts from YouTube videos.

-   **Enhanced Context Handling**: Uses message replies and context to understand user's intent better.

## Features

-   **User Access Control:** Only allows access to users specified in the configuration.

-   **Chat History:** Maintains chat history for each user, allowing for contextual conversations.

-   **File Handling:** Processes various file types, including images, documents, and audio files.

-   **Instagram Downloader:** Downloads Instagram posts and reels, with options to send as compressed media or files.

-   **YouTube Downloader:** Downloads audio from YouTube videos as MP3 files, including metadata and cover art.

-   **Web to Markdown Converter:** Converts webpages into markdown format for easy reading and sharing.

-   **Error Handling:** Includes robust error handling and retry mechanisms for network issues and API rate limits.

-   **Configuration:** Utilizes a configuration file for easy customization of bot settings, including allowed users, AI model, and safety settings.

-   **User Logging:** Logs user data including first and last seen dates, chat type, and more.

-  **Config Editor:** Built-in web based configuration editor to modify the bot settings and system prompts.

-  **Markdown Toggle**: Allows users to toggle between plain text and rendered markdown for bot responses.

-  **Enhanced Reply Context**: The bot can now understand and process replies to previous messages, including images, documents, and audio.

-  **Improved Media Handling**: Improved handling for audio and voice messages, including metadata and file sizes.

-  **Token Management**: Manages chat history token count to prevent exceeding API limits by removing the oldest messages.

-  **Detailed Reasoning**: The `/think` command provides detailed reasoning for the AI bot's responses.

## Commands

-   `/start`: Starts the bot and provides a welcome message.

-   `/help`: Displays a list of available commands.

-   `/clear`: Clears the chat history for the current user.

-   `/insta `: Downloads Instagram media from the provided URL as compressed media.

-   `/instaFile `: Downloads Instagram media from the provided URL as uncompressed files.

-   `/ytb2mp3 `: Downloads the audio from a YouTube video as an MP3 file.

-   `/web2md `: Converts a webpage to a markdown file.

-   `/ytb2transcript`: Extracts the transcript from a YouTube video by letting user to choose the language and the transcript format (with/without timestamps)

-   `/jailbreak` Inject jailbreak prompt into the current chat.

-   `/think `:  Provides a detailed reasoning for the AI bot's response to the query.

    > The Think command is a standalone function and not related to the main model or context and the generated responses are not saved in the context history.

## Setup

0.  **Recommended:**

    - Recommended to use Python version 3.11

    ```bash

    apt update

    apt install python3.11

    ```    

    

1.  **Initial Setup:**

    - Create and enter the virtual environment

    ```bash

    python -m venv venv

    source venv/bin/activitie

    ```

    - Install the requirements

    ```bash

    pip install -r requirements.txt

    ```

2.  **Install ffmpeg:**

    - Ensure you have `ffmpeg` installed on your system. This is required for YouTube to MP3 conversion.

3.  **Environment Variables:**

    -   Set the `TELEGRAM_TOKEN_KEY` environment variable with your Telegram bot token.

    -   Set the `GEMINI_API_KEY` environment variable with your Gemini API key.

4. **First Time Run:**

    -   Execute the `run.sh` script:

        ```bash

        ./run.sh

        ```

        This will start the bot and the configuration editor, allowing you to set up the bot's configuration.

        The config editor will be available at `http://127.0.0.1:8080`.

        After configuring the bot, you can close the editor.

5.  **Run the Bot:**

    -   After the first time run, you can start the bot directly by running:

        ```bash

        python main.py

        ```

        

   

**Note:** After the first run using `run.sh`, you can run the bot directly with `python main.py`. The `run.sh` script is mainly for creating the initial config file.

## License

This project is licensed under the [MIT](LICENSE) license.