OpenRouter Claw Code

Better Harness Tools, now working with any model from OpenRouter instead of Anthropic models only.

Overview

This is a fork of the project https://github.com/instructkr/claw-code/.

Project OpenRouter Claw Code aims to create a highly capable CLI agent harness. We have successfully updated this project to work with any model from OpenRouter, rather than being restricted to Anthropic models.

Currently, the repository contains two parallel tracks:

1. The Functional Agent (Rust): A fully working, high-performance CLI application that executes tools, maintains context, and talks to OpenRouter.
2. The Porting Workspace (Python): A clean-room Python rewrite currently in progress that captures the architectural patterns of the agent harness.

---

🚀 Getting Started: Using the Functional Agent (Rust)

To actually run the AI agent, interact with your codebase, and use OpenRouter models, you will use the Rust implementation. You will need Rust installed on your system.

1. Set your OpenRouter API Key

The agent authenticates using your OpenRouter API key. Export it in your terminal:

export OPENROUTER_API_KEY="sk-or-v1-..."

# Optional: If you are using a custom proxy endpoint
# export OPENROUTER_BASE_URL="https://your-custom-proxy.com/api"

2. Start the Interactive REPL

Navigate to the rust directory and run the application:

cd rust/
cargo run --release

This will drop you into the interactive claw REPL where you can type prompts or use slash commands (like /help or /status).

3. Using Specific OpenRouter Models

By default, the agent tries to use an Opus alias, but you can pass any OpenRouter model string or use built-in aliases (like gpt4o, deepseek, sonnet, haiku):

# Using a built-in alias:
cargo run --release -- --model google/gemini-3-flash-preview

# Using an exact OpenRouter model ID:
cargo run --release -- --model meta-llama/llama-3-70b-instruct

4. One-Shot Prompt Mode

To run a single command and exit without entering the REPL:

cargo run --release -- prompt "Analyze the files in this directory and summarize them"

---

🐍 Development: Python Porting Workspace

The main source tree (src/) contains the active Python porting workspace. The current Python workspace is not yet a complete one-to-one replacement for the Rust system, but it is the primary implementation surface for the rewrite.

Repository Layout

.
├── rust/                               # Functional, high-performance Rust CLI agent
├── src/                                # Python porting workspace (WIP)
│   ├── __init__.py
│   ├── commands.py
│   ├── main.py
│   ├── models.py
│   ├── port_manifest.py
│   ├── query_engine.py
│   ├── task.py
│   └── tools.py
├── tests/                              # Python verification
└── README.md

Python Workspace Overview

The new Python src/ tree currently provides:

• port_manifest.py — summarizes the current Python workspace structure
• models.py — dataclasses for subsystems, modules, and backlog state
• commands.py — Python-side command port metadata
• tools.py — Python-side tool port metadata
• query_engine.py — renders a Python porting summary from the active workspace
• main.py — a CLI entrypoint for manifest and summary output

Python Quickstart

If you want to explore the Python rewrite codebase, you can run the following commands from the root of the repository (requires Python 3):

# Render the Python porting summary:
python3 -m src.main summary

# Print the current Python workspace manifest:
python3 -m src.main manifest

# List the current Python modules:
python3 -m src.main subsystems --limit 16

# Inspect mirrored command/tool inventories:
python3 -m src.main commands --limit 10
python3 -m src.main tools --limit 10

# Run verification tests:
python3 -m unittest discover -s tests -v