Claw Code

Plugin-Based Tool System

Offers 19 permission-gated tools for file I/O, shell execution, Git operations, and web scraping, each independently sandboxed with configurable access controls.

Autonomous Agent Loop

Provides a terminal-native agent that reads entire codebases, edits files, executes commands, runs tests, handles Git, and iterates autonomously until task completion.

Multi-Agent Orchestration

Supports spawning sub-agents, internally referred to as "swarms," to parallelize complex tasks, enabling decomposition into independently executable subtasks with shared memory access.

LLM API Client

Delivers a provider-agnostic API client with automatic retry, SSE streaming, OAuth authentication, and token usage tracking with cost estimation, supporting multiple LLM providers.

Session & Memory Management

Implements a multi-layer memory system with session persistence, transcript compaction, and context discovery, ensuring persistent knowledge across conversations with automatic cleanup.

Query Engine

Serves as the central intelligence, managing all LLM API calls, response streaming, caching strategies, and multi-step orchestration with provider-agnostic design and configurable turn limits.

Rust Performance Core

Features a 6-crate Rust workspace with 16 runtime modules for performance-critical paths, zero-dependency JSON parser, OAuth PKCE flow, and syntax-highlighted terminal rendering.

MCP Integration

Provides full Model Context Protocol support with 6 transport types—Stdio, SSE, HTTP, WebSocket, SDK, and ClaudeAiProxy—for connecting to external tool servers with automatic name normalization and OAuth authentication.

Slash Commands

Offers 15 interactive commands (/compact, /model, /permissions, /cost, /session) with resume support, command graph categorization, and REPL integration for comprehensive session control.

Permission System

Implements three permission modes with a per-tool policy engine, deny lists, and interactive prompting, ensuring granular access control and security across all agent operations.

• Developers: Build custom AI coding agents using Claw Code's plugin-based tool system and extensible architecture.
• Open-source contributors: Enhance the Claw Code framework by contributing to its modular Python and Rust codebase.
• Researchers: Study the clean-room reimplementation of Claude Code's architecture for AI agent development insights.
• Enterprises: Deploy self-hosted AI coding agents with Claw Code's multi-LLM provider support and permission controls.
• Students: Learn about AI agent architecture and implementation by exploring Claw Code's open-source codebase.

What is Claw Code?

Claw Code is an open-source AI coding agent framework built in Rust and Python. It is a clean-room rewrite of the Claude Code agent harness architecture, created by Sigrid Jin after the March 2026 source code leak. The project reimplements core architectural patterns — including the tool system, query engine, multi-agent orchestration, and memory management — without copying any proprietary source code.

Does Claw Code contain Anthropic's proprietary code or model weights?

No. Claw Code is a clean-room implementation written entirely from scratch. Independent code audits confirm that the project contains no Anthropic proprietary source code, no model weights, no API keys, and no user data. The implementation is architecturally informed but legally independent.

How is Claw Code different from Claude Code?

Claude Code is Anthropic's official proprietary CLI agent written in TypeScript, requiring a Claude subscription. Claw Code is an open-source alternative written in Python and Rust, supporting multiple LLM providers (Claude, OpenAI, local models). Claw Code offers a modular, extensible architecture that developers can customize and self-host. See the full comparison.

What languages and technologies does Claw Code use?

The codebase is composed of Rust (72.9%) and Python (27.1%). Python handles the agent orchestration layer, command parsing, and LLM integration. Rust implements performance-critical runtime paths, with an active migration branch targeting a fully native runtime for maximum performance and memory safety.

Who created Claw Code?

Claw Code was created by Sigrid Jin (@sigridjineth), a developer profiled by the Wall Street Journal as one of the world's most active Claude Code users, having processed over 25 billion tokens. The project is maintained under the instructkr organization on GitHub and welcomes community contributions.

Is Claw Code safe to use?

Claw Code itself is open source and auditable. However, users should exercise caution with the broader ecosystem: a supply-chain attack on March 31, 2026 affected npm-based Claude Code installations. Always install from official sources, verify dependency integrity, and review lockfiles for unauthorized packages.

• Visit the Claw Code GitHub repository to clone the project and begin setup.
• Install Python dependencies using pip install -r requirements.txt to prepare the environment.
• Run the agent by executing python src/main.py to start interacting with the CLI.
• Explore the modular architecture, which includes tools, commands, and a query engine for agent orchestration.
• Customize the framework by modifying tool permissions, adding new commands, or integrating alternative LLM providers.
• Refer to the official documentation for advanced configuration, such as MCP integration or Rust runtime optimization.
• Join the community discussions on GitHub to contribute improvements or seek support for deployment and usage.