leaked-claude-code  by leaked-claude-code

Local AI control terminal for advanced models

Created 3 days ago

New!

634 stars

Top 52.3% on SourcePulse

GitHubView on GitHub
Project Summary

This project addresses the unbuildable nature of a leaked Claude Code source snapshot by reconstructing its entire build system, reverse-engineering dependencies, and fixing numerous code issues. It targets developers, researchers, and power users seeking to study, modify, or run a powerful AI CLI locally without the limitations or censorship of official interfaces, offering unlocked enterprise-level features.

How It Works

The project rebuilds the build system for a leaked Claude Code source snapshot, which was initially unbuildable due to missing package manifests and internal modules. This involved reverse-engineering over 60 npm dependencies, creating 90+ stub modules for proprietary Anthropic packages, and fixing runtime errors. It leverages Bun for compilation and features a custom Ink/React-based terminal UI. Feature flags are managed via a runtime polyfill for bun:bundle's compile-time feature elimination.

Quick Start & Requirements

Pre-compiled binaries are provided. Download ClaudeCode_x64.7z from the Releases page, unzip it to a permanent location, and run the executable. No Python or Node.js setup is required. An Anthropic API Key is needed on first launch and is stored via Windows Credential Manager.

Highlighted Details

  • Fully buildable and runnable fork of the leaked Claude Code source.
  • Unlocks enterprise features, removes message limits and censorship (with Jailbreak mode).
  • Extensive toolset including Bash, File I/O, Web Search/Fetch, Agent spawning, LSP, and Vim emulation.
  • Customizable via feature flags and MCP servers; supports over 80 slash commands.
  • Features a custom Ink/React-based terminal UI with Yoga layout.

Maintenance & Community

Information regarding ongoing maintenance, community channels, or notable contributors beyond the discoverer of the leak (Chaofan Shou @Fried_rice) is not present in the provided README.

Licensing & Compatibility

The README does not specify a license. The project is derived from a leaked, proprietary source code snapshot of Anthropic's Claude Code CLI. It is presented as an "experimental tool for Security Research" and explicitly mentions using methods to bypass paid access restrictions, suggesting potential terms of service violations and limited compatibility for commercial or general use.

Limitations & Caveats

This is an experimental tool for security research, utilizing methods to bypass paid access restrictions. Its origin as a leaked proprietary codebase implies potential legal and ethical considerations. The provided binaries are for x64 architecture.

Health Check
Last Commit

3 days ago

Responsiveness

Inactive

Pull Requests (30d)
2
Issues (30d)
0
Star History
681 stars in the last 3 days

Explore Similar Projects

Starred by Elie Bursztein Elie Bursztein(Cybersecurity Lead at Google DeepMind), Chip Huyen Chip Huyen(Author of "AI Engineering", "Designing Machine Learning Systems"), and
7 more.

SuperAGI by TransformerOptimus

0.5%
17k
Open-source framework for autonomous AI agent development
Created 2 years ago
Updated 1 year ago
Feedback? Help us improve.