@jhinresh/game-maiat-plugin
v0.9.0
Published
Maiat Trust Score plugin for the GAME SDK — lets your Virtuals agent verify trust before swapping or transacting
Maintainers
Readme
@virtuals-protocol/game-maiat-plugin
Trust score verification for GAME agents — powered by Maiat Protocol
Give your Virtuals agent the ability to check trust before it acts. Maiat Protocol provides on-chain verified trust scores for tokens, DeFi protocols, and AI agents on Base.
Why
AI agents executing swaps autonomously can be exploited by:
- Rug-pull tokens with no history
- Honeypot contracts
- Low-liquidity traps
- Unverified counterparties
Maiat trust scores are computed from verified reviews, on-chain activity, and AI analysis — stored in a TrustScoreOracle contract on Base.
Install
npm install @virtuals-protocol/game-maiat-pluginQuick Start
import { GameAgent, GameWorkerService } from "@virtuals-protocol/game";
import MaiatTrustPlugin from "@virtuals-protocol/game-maiat-plugin";
const maiatPlugin = new MaiatTrustPlugin({
minScore: 3.0, // reject tokens with score < 3.0/10
chain: "base",
});
const agent = new GameAgent(process.env.GAME_API_KEY!, {
name: "Trust-Gated Agent",
goal: "Execute swaps only for tokens with Maiat trust score ≥ 3.0",
workers: [new GameWorkerService(maiatPlugin.getWorker())],
});
await agent.init();
await agent.step({ verbose: true });Functions
check_trust_score
Query the trust score for any address or project name.
| Arg | Type | Description |
|-----|------|-------------|
| identifier | string | 0x... address or project name (e.g. "Uniswap", "AIXBT") |
| chain | string | Blockchain (default: "base") |
Returns:
{
"address": "0x...",
"score": 8.7,
"risk": "low",
"type": "DeFi",
"flags": [],
"safe": true,
"summary": "Trust score for Uniswap: 8.7/10 | Risk: LOW | Safe: YES"
}gate_swap
Check whether a swap is safe before executing. Checks both sides.
| Arg | Type | Description |
|-----|------|-------------|
| token_in | string | Token being sold |
| token_out | string | Token being bought |
| min_score | number | Minimum score to pass (default: 3.0) |
Returns: APPROVED or REJECTED with reasons.
batch_check_trust
Score multiple addresses at once, sorted by trust.
| Arg | Type | Description |
|-----|------|-------------|
| identifiers | string | Comma-separated addresses or names (max 10) |
Configuration
new MaiatTrustPlugin({
apiUrl: "https://app.maiat.io", // default
apiKey: "your-api-key", // optional
minScore: 3.0, // 0–10 scale
chain: "base", // base | bnb | solana
})Score Scale
| Score | Risk | Meaning | |-------|------|---------| | 7–10 | 🟢 Low | Trusted, well-reviewed | | 4–6.9 | 🟡 Medium | Use with caution | | 0–3.9 | 🔴 High | Risky, avoid |
Links
- 🌐 app.maiat.io
- 📦 GitHub
- 🔗 Oracle contract (Base Sepolia):
0xF662902ca227BabA3a4d11A1Bc58073e0B0d1139 - 📖 API docs
License
MIT
