Developer SDK Preview
Add real-time ad safety detection to your website or app.
The AEGIS SDK is designed to help developers detect harmful or deceptive advertising content, apply configurable safety policies, and explain moderation decisions directly inside their products.
SDK Overview
Designed for minimal setup and maximum control.
The JavaScript SDK will allow web developers to integrate AEGIS into websites and applications with minimal setup. The SDK can observe rendered content, detect ad-like elements, analyze safety signals, and apply policy actions such as labeling, blurring, or blocking.
npm install @aegis-safety/js-sdk
# or
pnpm add @aegis-safety/js-sdkimport { Aegis } from "@aegis-safety/js-sdk";
const aegis = new Aegis({
policy: "standard",
enforcement: "label-and-blur",
explainability: true,
});
// Scan the current page for ad-like content
aegis.scan();Developer Capabilities
Everything you need to build safer products.
DOM scanning
Identify ad-like and promotional content on rendered pages in real time.
Policy hooks
Apply custom safety policies based on your product's audience and requirements.
Risk scoring
Receive structured risk scores, categories, and confidence levels for each detection.
Moderation actions
Allow, label, blur, or block unsafe content with a single enforcement call.
Explainability
Surface understandable moderation reasons directly to users or in your UI.
Reporting hooks
Prepare for future analytics, audit trails, and safety reporting integrations.
Architecture
How the SDK fits into your stack.
Website / App
Renders content from ad networks and third-party sources
AEGIS JS SDK
Lightweight integration layer installed by developers
Ad Detection Layer
Identifies ad-like and promotional content in the DOM
Core Policy Engine
Evaluates content against configured safety rules
AI + Rules + Reputation
Multi-layer signals produce a structured risk score
Risk Score + Explanation
Structured output with category and confidence
Enforcement Action
Allow, label, blur, or block applied to the content
Policy Configuration
Tune safety behavior for your audience.
Developers should be able to configure enforcement sensitivity per category — applying strict defaults for children, configurable thresholds for general audiences, and lenient settings for enterprise research tools.
const result = await aegis.analyze({
text: "Limited-time investment opportunity. Guaranteed returns.",
url: "https://example-promo.test",
});
if (result.riskScore > 0.8) {
aegis.enforce("block", {
reason: result.explanation,
});
}const aegis = new Aegis({
policy: {
childSafety: "strict",
scams: "block",
adultContent: "block",
suspiciousFinancialClaims: "label",
unknownDomains: "blur",
},
});Want early access to the AEGIS SDK?
Join the developer waitlist to follow SDK development, test early builds, or contribute to the open-source core.