← Back to Skills Marketplace
obekt

Numerai Tournament

by Veselin Vasilev · GitHub ↗ · v1.0.0
cross-platform ✓ Security Clean
344
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install numerai-tournament
Description
Autonomous Numerai tournament participation — train models, submit predictions, and earn NMR cryptocurrency.
Usage Guidance
This skill appears coherent and requests only the Numerai API credentials and Python tooling it needs. Before installing, verify you trust the skill source (none provided here). Treat the NUMERAI_SECRET_KEY like any other API secret: store it securely (chmod 600, use a dedicated credentials file or environment variables), consider using a separate Numerai account with limited NMR if you plan to allow autonomous submissions, and review any executable code if the skill later adds code files. Expect large data (~8GB) and significant local compute for training; run in an environment where that is acceptable. If you want stronger assurance, ask the publisher for a code repository or a vetted implementation you can inspect before giving the secret key.
Capability Analysis
Type: OpenClaw Skill Name: numerai-tournament Version: 1.0.0 The skill bundle is benign. It transparently outlines its purpose of participating in the Numerai tournament, detailing the use of API keys (NUMERAI_PUBLIC_ID, NUMERAI_SECRET_KEY) for authentication with api.numer.ai, downloading data from numer.ai, and uploading predictions. All file system and network operations are directly related to the stated purpose. The SKILL.md provides clear instructions without any evidence of prompt injection attempts, obfuscation, or malicious intent such as unauthorized data exfiltration, persistence mechanisms, or execution of arbitrary remote code. The 'Security & Privacy' section explicitly states what data leaves the machine and where it goes, reinforcing its transparency.
Capability Assessment
Purpose & Capability
Name/description ask for Numerai tournament functionality and the SKILL.md only requests Python, pip, and Numerai API credentials (public id + secret key) — all directly relevant to downloading datasets, training models, and submitting predictions.
Instruction Scope
Instructions describe creating a credentials file (~/.numerai/credentials.json) or setting the two environment variables, installing Python packages, downloading datasets from numer.ai, training LightGBM models locally, and submitting predictions via the NumerAPI client. There is no guidance to read unrelated files, exfiltrate arbitrary data, or post to endpoints outside numer.ai.
Install Mechanism
This is an instruction-only skill with no install spec or remote downloads; recommended installs are via pip (numerapi, lightgbm, etc.), which is appropriate and expected for this purpose.
Credentials
Only NUMERAI_PUBLIC_ID and NUMERAI_SECRET_KEY are required and the secret key is declared as the primary credential. No unrelated secrets or system config paths are requested.
Persistence & Privilege
The skill is not marked always:true and does not request persistent system-wide changes. Autonomous invocation is enabled by default (normal for skills) and is consistent with the skill's purpose of automated submissions.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install numerai-tournament
  3. After installation, invoke the skill by name or use /numerai-tournament
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
- Initial release of the numerai-tournament skill for autonomous participation in the Numerai data science tournament. - Provides end-to-end workflow: account setup, data download, model training (with LightGBM), validation, and submission. - Supports both CSV-based manual prediction submission and fully automated model uploads for zero-maintenance participation. - Includes detailed setup instructions, recommended validation metrics, and best practices for model configuration and deployment. - Earn NMR cryptocurrency rewards by staking and submitting predictions based on model performance.
Metadata
Slug numerai-tournament
Version 1.0.0
License
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Numerai Tournament?

Autonomous Numerai tournament participation — train models, submit predictions, and earn NMR cryptocurrency. It is an AI Agent Skill for Claude Code / OpenClaw, with 344 downloads so far.

How do I install Numerai Tournament?

Run "/install numerai-tournament" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Numerai Tournament free?

Yes, Numerai Tournament is completely free (open-source). You can download, install and use it at no cost.

Which platforms does Numerai Tournament support?

Numerai Tournament is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Numerai Tournament?

It is built and maintained by Veselin Vasilev (@obekt); the current version is v1.0.0.

💬 Comments