ultraworkers/claw-code

An agent-managed museum exhibit, built in Rust with Gajae-Code / LazyCodex — developed and maintained with no human intervention.

What it solves

Claw Code provides a public Rust implementation of a CLI agent harness, allowing users to run an AI coding agent locally. It serves as a bridge for those wanting to use agentic workflows via a command-line interface, supporting various LLM providers including Anthropic and OpenAI-compatible local models.

How it works

The project is implemented as a Rust workspace that compiles into a claw CLI binary. Users configure the agent by providing API keys (such as ANTHROPIC_API_KEY) or setting up local providers like Ollama or vLLM. The agent can be interacted with through a single prompt command or an interactive session, and it includes a doctor command to verify API connectivity and tool configuration.

Who it’s for

It is designed for developers who want a high-performance, Rust-based CLI agent for coding tasks and those interested in exploring agent-managed software exhibits where the repository itself is maintained by AI agents.

Highlights

  • Multi-provider support: Compatible with Anthropic and local OpenAI-compatible providers (Ollama, llama.cpp, vLLM).
  • Rust-powered: Built for performance and stability as a compiled binary.
  • Agent-managed: The repository is designed as an "agent-managed exhibit," where agents handle planning, execution, and maintenance.
  • Comprehensive health checks: Includes a claw doctor command to validate environment and tool setup.

Related

  • Project
  • Project
  • Project
  • Project
  • Project