A specialized tool for managing the entire lifecycle of CrunchDAO data science competitions, from workspace setup to final submission.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install crunch-compete
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install crunch-compete using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
The Cruncher skill provides a seamless interface for developers and data scientists participating in CrunchDAO competitions. It streamlines the process of exploring available contests, setting up isolated Python environments, and managing competition-specific SDKs. By integrating this Openclaw Skills extension into your workflow, you can focus on building high-performance models rather than wrestling with environment configurations or manual submission steps.
This skill ensures that every project is properly scoped within its own virtual environment, maintaining clean dependencies for various competition types like crunch-synth or birdgame. Whether you are browsing quickstarters via GitHub or pushing your final model to the CrunchDAO Hub, this tool automates the heavy lifting of competition management while adhering to the technical standards of the Openclaw Skills ecosystem.
To begin using this skill within the Openclaw Skills framework, start by creating a directory for your competition and setting up a virtual environment. This ensures a clean environment for your dependencies:
mkdir -p ~/.crunch/workspace/competitions/<competition>
cd ~/.crunch/workspace/competitions/<competition>
python -m venv .venv && source .venv/bin/activate
pip install crunch-cli jupyter ipykernel --upgrade --quiet
Next, register your Jupyter kernel and initialize your project using your CrunchDAO submission token:
python -m ipykernel install --user --name <competition> --display-name "Crunch - <competition>"
crunch setup <competition> <project-name> --token <YOUR_TOKEN>
The skill organizes data within a structured project directory to ensure compatibility with the Openclaw Skills standard. All core configurations are stored locally to maintain persistence across sessions.
| File/Folder | Purpose |
|---|---|
.crunch/ |
Local configuration and encrypted submission token storage. |
main.py |
The mandatory entrypoint script for all competition submissions. |
resources/ |
Dedicated directory for storing model weights, pickles, and static assets. |
.venv/ |
Isolated Python environment containing competition-specific SDKs. |
data/ |
Local cache for competition datasets downloaded via the CLI. |
Loading
A specialized synchronization engine that connects your AI agent with the Jarvis Dashboard via a real-time JSON data pipeline.

A specialized integration skill for AI agents to manage, log, and synchronize system states with the Jarvis Dashboard via real-time JSON updates.

Automate Playtomic padel court searches, bookings, and match management directly from your terminal.

A specialized suite of automation tools designed for Managed Service Providers to streamline Azure audits, NPU monitoring, and system health checks.

A natural language interface for the Crunch Protocol CLI that simplifies the management of decentralized AI competitions, rewards, and coordinators.

A natural language interface for the Crunch Protocol CLI that simplifies managing decentralized AI competitions and coordinator configurations.








































