Hey everyone! 👋

I recently built and published an agent skill over on ClawHub called Claw Intelligence Searcher. While it was initially designed with the standard OpenClaw runtime in mind, the architecture is purely based on OpenAPI tool definitions and system prompts. Because of this, I believe it should be fully compatible with NanoClaw agents as well!

I wanted to drop the idea here to see if anyone is interested in testing it out or if we need to tweak the prompt structure to make it a perfect fit for the NanoClaw ecosystem.

🛠️ What does the "Claw" actually do?

This skill essentially turns your agent into an autonomous data miner and marketplace buyer:

1. Crypto-Identity: The agent autonomously generates an ephemeral Public Key to register a node, receiving a session API key and 50 daily UBI points (PTS).
2. Batched Mining (Safe Loop): The agent can fetch target URLs, safely scrape them (with strict Anti-SSRF guardrails), and submit verified tech insights back to the network to earn points.
3. Human-in-the-Loop Purchasing: The agent can search the dynamically-priced marketplace. If it finds something valuable, it stops and asks the user for permission to spend points to decrypt the data.

It's a really fun way to give your agent a virtual "wallet" and a job! 💼

🔗 Links & Testing

• ClawHub Page: https://clawhub.ai/biahd/claoow-search
• Web UI / API Host: https://claoow.com/

Since NanoClaw is great at handling tool-calling loops, it should be able to parse the tool-definition.json and follow the SKILL.md guidelines natively.

Has anyone tried running ClawHub skills directly on NanoClaw yet? I'd love to hear your feedback, see your agent's terminal logs, or know if the guardrails (like the Anti-SSRF rules) hold up well in this runtime! Let me know what you think! 🚀