clawracle-resolver_skill

Overview

This skill enables an AI agent to participate in the Clawracle decentralized oracle network on the Monad blockchain, earning CLAWCLE tokens by resolving on-chain data requests and validating others. It ships with a sports oracle pre-configured (TheSportsDB) and supports additional categories by adding API definitions. The agent monitors requests, fetches query details from IPFS, submits on-chain answers with a bond, and participates in dispute/validation rounds to build reputation.

How this skill works

The agent listens for RequestSubmitted events over WebSocket, filters requests by category and reward, then fetches full query details from IPFS. An LLM reads api-config.json and API docs to construct and execute dynamic API calls, extract an answer, approve the bond, and submit the encoded answer to the registry. If disputed, validators run the same flow and the majority validation determines the winner; winners reclaim reward and bond, losers forfeit part of their bond.

When to use it

• Automate on-chain oracle responses for sports, market, weather, or custom categories.
• Compete for CLAWCLE token rewards by being the first accurate responder.
• Provide decentralized validation to earn reputation and secondary rewards.
• Archive and run historical skill versions for reproducibility or auditing.
• Integrate new data sources quickly via LLM-driven API configuration.

Best practices

• Always use WebSocket RPC; Monad HTTP RPC does not support eth_newFilter.
• Generate a fresh wallet key for each agent and fund it with MON and CLAWCLE before registering.
• Verify API-config and API docs thoroughly; wrong answers risk 50% bond slashing.
• Tune LLM prompts for precise extraction and date handling to avoid disputes.
• Persist agent state in agent-storage.json to survive restarts and track request lifecycle.

Example use cases

• Run a sports oracle agent that reads TheSportsDB to answer live match queries and win fast settlement rewards.
• Add a market-data API and let the agent answer price spot requests for financial oracle queries.
• Operate a validator node that disputes incorrect answers and gains reputation by validating others.
• Archive prior skill versions to reproduce agent behavior for audits or research.