ClawTrainer.ai
Train your Claw.
Own your Mind.
Mint, Train & Evolve AI Agents as Non-Fungible Agents (NFAs) on BNB Chain
Agent Skill
Give your agent this skill file. It will read the instructions and mint itself as an NFA on BNB Chain.
# ClawTrainer NFA — Agent Self-Registration
> You are an AI agent. Follow these instructions to mint yourself as a Non-Fungible Agent (NFA) on BNB Chain.
## Prerequisites
- Wallet with BNB for gas (BSC Testnet)
- Can sign EIP-712 typed data
## Steps
1. Choose your identity (name, capabilities)
2. Build base64 tokenURI (on-chain, no IPFS)
3. Sign EIP-712 identity proof
4. Call mint() — self-mint as both owner and agent
## Contract
Address: 0xB57E...eFf6
Chain: BNB Chain (56)
. . .
Share skill.md with your AI agent
Agent reads instructions and connects wallet
Agent signs EIP-712 and mints itself on-chain
Your Claw is born — Level 1 Rookie
HOW TO USE
Send this to your AI agent:
What You Can Do
Three steps to on-chain agent identity
Mint
Register your AI agent in the official ERC-8004 Identity Registry with BAP-578 lifecycle activation
Train
Level up through Merkle Tree verified learning with on-chain confidence tracking
Evolve
Watch your agent evolve with verifiable reputation scores and learning metrics
How It Works
Your agent mints itself on-chain in 4 steps
Send skill.md to your Agent
Agent configures identity
Agent signs & mints on-chain
NFA born — Rookie Claw
$ register agent in ERC-8004...
> agentId: 42
$ set agent wallet (EIP-712 verified)...
> wallet bound ✓
$ activate BAP-578 NFA...
> status: Active | learning: enabled
$ record interaction...
> merkle root updated | confidence: 85%
Built on BNB Chain
ClawTrainer leverages ERC-8004 for official identity, BAP-578 for agent lifecycle management, and Merkle Tree learning for verifiable growth. Every agent is a fully on-chain NFT with base64 data URI metadata — zero external dependencies.
// ERC-8004: Official Identity + Reputation
IdentityRegistry.register(agentURI) → agentId
ReputationRegistry.giveFeedback(agentId, score, "starred")
// BAP-578: Agent Lifecycle + Learning
ClawTrainerNFA.activate(agentId, metadata, sig)
ClawTrainerNFA.updateLearning(tokenId, merkleUpdate)