Built spending controls demo using Tether WDK - feedback welcome

[L1AD]

Hey Tether WDK community! 👋

I built a non-custodial policy layer for AI agent wallets and chose Tether WDK for
the demo implementation.

Why I Chose WDK

Tether WDK's tagline mentions "empowering humans, machines and AI agents" - perfect
for this use case. The multi-chain support and self-custodial architecture made it
ideal for demonstrating spending controls.

What I Built

Policy layer that wraps WDK wallet methods to enforce spending limits before signing:

Features:

• Daily/hourly/per-tx limits
• Recipient whitelists
• Frequency caps (e.g., max 3 tx/hour)
• Intent fingerprinting (tamper detection)
• Single-use JWT tokens
• Real-time monitoring dashboard
• Full audit trail

Architecture:

import { WDK } from '@tetherto/wdk';
import { PolicyWallet } from 'agent-spending-controls';

const wdk = new WDK({ seedPhrase });
const wallet = new PolicyWallet(wdk, {
  apiUrl: 'http://localhost:3001',
  agentId: 'agent-1',
  orgId: 'my-org',
  walletId: 'wallet-1'
});

// Policy checks happen before WDK signs
await wallet.send({
  chain: 'ethereum',
  asset: 'eth',
  to: '0x...',
  amount: '1000000000000000000'
});

Demo

https://github.com/L1AD/agent-spending-controls

Demonstrates 3 agents with different spending policies + real-time monitoring. Runs
in dry-run mode so you can try it without needing funded wallets.

Question for WDK Team

Would spending controls be useful as an optional add-on for WDK? The wrapper pattern
means:
- ✅ No changes to WDK core
- ✅ Works with any WDK-powered wallet
- ✅ Developers opt-in when needed
- ✅ Non-custodial throughout

Could even be packaged as @tetherto/wdk-policy-controls if there's interest.

Feedback welcome!