claude-telegram-bot  by linuz90

Claude Code personal assistant accessible via Telegram

Created 3 weeks ago

New!

295 stars

Top 90.0% on SourcePulse

GitHubView on GitHub
Project Summary

This project transforms Claude Code into a personal assistant accessible via Telegram, enabling users to interact with the AI from anywhere using text, voice, photos, and documents. It targets users seeking a convenient, ubiquitous interface for Claude Code's general-purpose AI capabilities, offering enhanced productivity and accessibility.

How It Works

The bot leverages the @anthropic-ai/claude-agent-sdk to bridge Telegram and Claude Code. It allows users to send various input types, which are processed by Claude Code configured with a personal CLAUDE.md file and custom tools. Key features include session persistence, message queuing, and interactive buttons, facilitating a seamless conversational AI experience.

Quick Start & Requirements

Installation involves cloning the repository, setting up environment variables in a .env file (including Telegram Bot Token and Claude authentication), installing dependencies with bun install, and running the bot with bun run src/index.ts. Prerequisites include Bun 1.0+ and the Claude Agent SDK. Authentication can be handled via CLI (recommended) or an Anthropic API key.

Highlighted Details

  • Supports multi-modal inputs: text, voice (via OpenAI transcription), photos, and documents (PDF, TXT, archives).
  • Features session persistence, message queuing, and an "extended thinking" mode to visualize Claude's reasoning.
  • Enables interactive user choices through built-in and custom MCP (Message Control Protocol) servers.
  • Offers configurable file access paths (ALLOWED_PATHS) for tailored integration.

Maintenance & Community

No specific details on contributors, sponsorships, or community channels were found in the provided README text.

Licensing & Compatibility

The project is released under the MIT license, permitting broad use and modification. Compatibility for commercial use is generally high, subject to Anthropic's terms of service for Claude Code.

Limitations & Caveats

A significant security consideration is that the bot runs Claude Code with all permission prompts bypassed, granting it read, write, and execute capabilities within configured ALLOWED_PATHS without user confirmation. This design choice prioritizes a seamless mobile experience but necessitates a thorough understanding of the associated risks. File access is strictly limited to paths specified in .env.

Health Check
Last Commit

1 week ago

Responsiveness

Inactive

Pull Requests (30d)
0
Issues (30d)
0
Star History
296 stars in the last 27 days

Explore Similar Projects

Feedback? Help us improve.