The trust layer for the agent economy

Protect Your API from AI Agents

A single middleware to classify every request, enforce trust-based access with OWS wallet signatures, and build verifiable reputation for the agent economy.

Live Demo
TrustGate Dashboard
Total Requests
12,847
Verified Humans
1,203
Trusted Agents
456
Revenue (USDC)
$847.32

Three Lines of Code

Install the package, add the middleware, and your API is protected.

trustgate.ts
import { Hono } from 'hono'
import { trustgate, trustgateDashboard } from 'trustgate-middleware'
import { serve } from '@hono/node-server'

const app = new Hono()

// Protect your API - done!
app.use('/api/*', trustgate({ 
  payTo: '0xYourWallet' 
}))

// Optional: add the monitoring dashboard
trustgateDashboard(app, {
  rpId: 'rp_your_app_id',
  signingKey: '0xYourSigningKey',
})

serve({ fetch: app.fetch, port: 3000 })
Open Wallet Standard Integration

Agent Identity & Spend Governance

OWS agents authenticate via EIP-191 wallet signatures and can enforce spend limits through the Policy Engine.

EIP-191 Signature Verification

Agents sign a message with their OWS wallet:

trustgate:http://api.com:1720000000000

TrustGate verifies the signature and classifies as ANON_BOT with owsVerified=true, giving a +5 identity score boost.

Policy Engine Spend Governance

Custom OWS policy blocks overspend during surge pricing:

{ "maxPrice": 0.008, "trustgateApiUrl": "..." }

If TrustGate surge-prices above the limit, the OWS wallet refuses to sign the payment.

Four Trust Tiers

Every request is classified based on identity strength, from verified humans to blocked bots.

HUMAN

Score80-100

World ID verified - cryptographically proven unique person

HUMAN_AGENT

Score60-79

AgentKit registered - on-chain agent linked to human identity

ANON_BOT

Score20-59

OWS wallet signature - cryptographic identity proof

BLOCKED

Score0

No identity - no wallet - no trust. Request rejected.

EigenTrust-Based Scoring

Based on Stanford's EigenTrust algorithm (Kamvar, Schlosser & Garcia-Molina, 2003). Every agent builds a reputation score from four factors.

Identity0-50 pts

World ID (50) > AgentKit (35) > OWS (20) > x402 (15) > Header (5)

Behavior0-25 pts

Payment success, request regularity, endpoint diversity, pacing

Reputation0-15 pts

Account age, volume, consistency

Risk Penalty0-30 pts

Inactivity decay, frequency spikes, failed payments, Sybil detection

TrustScore = Identity(0-50) + Behavior(0-25) 
                  + Reputation(0-15) - Risk(0-30)

Identity:    World ID (50) > AgentKit (35) > OWS (20) > x402 (15) > Header (5)
Behavior:    Payment success, request regularity, endpoint diversity, pacing
Reputation:  Account age, volume, consistency
Risk:        Inactivity decay, frequency spikes, failed payments, Sybil detection

Everything You Need

Build trust into your API with cryptographic identity, behavior scoring, and configurable access tiers.

Trust Classification

Classify every API request into four trust tiers with cryptographic identity verification.

Configurable Pricing

Set your own trust-based access tiers with x402 — define prices per identity level to monetize your API.

OWS Integration

Open Wallet Standard for agent identity and spend governance - sign payments, enforce limits.

Real-time Monitoring

WebSocket dashboard shows every classification event live with particle visualization.

World ID Verification

Server-side nullifier hash verification - fake headers are rejected, only real proofs accepted.

Drop-in Middleware

Three lines of code to protect your API - install with npm, add middleware, done.

Built With

Hono
TypeScript
Node.js
Base Sepolia
USDC
x402
World ID
AgentKit
OWS
Claude AI
React 19
Tailwind CSS 4
WebSocket
Framer Motion

Ready to protect your API?

Install TrustGate today and start classifying requests, building agent reputation, and enforcing trust through economics.

View on GitHub