NeuralTrust | Platform for Agent Security.

Create an API key on the collector first. Set direction on every call (input before the model, output after). Defaults to input if omitted.

Coverage

Surface	Monitor	Block	Redact
LLM input	✅	⚠️	⚠️
LLM output	✅	⚠️	⚠️
Tool-level	⚠️	⚠️	⚠️

Ask — status is advisory: nothing prompts anyone unless your code does. Use it when the model call lives in your JavaScript or TypeScript code and the policy decision needs application context. Not as your only defence when you need a guarantee that does not depend on every developer remembering the check. ⚠️ TrustGuard returns the verdict, your code enforces it. Ignoring transformedPayload silently disables redaction: the call succeeds, the finding is recorded, and the unmasked prompt reaches the model. Limits. Coverage is per call site — see Node middleware to cover by configuration instead. Full comparison: Coverage.

npm install @neuraltrust/trustguard-sdk

import { TrustGuard } from "@neuraltrust/trustguard-sdk";

const client = new TrustGuard({
  baseUrl: "<your-trustguard-url>",
  apiKey: "<collector-api-key>",
});

const inbound = await client.guard({
  payload: { input: userInput },
  direction: "input",
  consumerId: user.id,
  sessionId: conversationId,
});
if (inbound.isBlocked) throw new Error("Blocked by TrustGuard (input)");
const prompt = inbound.transformedPayload ?? { input: userInput };

// … model call …

const outbound = await client.guard({
  payload: { input: modelCompletion },
  direction: "output",
  consumerId: user.id,
  sessionId: conversationId,
});
if (outbound.isBlocked) throw new Error("Blocked by TrustGuard (output)");

Introduction

Resources

Detector catalog

Integrations

Evaluate API

Node.js SDK

Coverage

​Coverage

Coverage