Sightenginejs

A fully‑typed TypeScript SDK for the SightEngine content‑moderation API, with streaming support, URL/file inputs, threshold presets, and helper utilities.🧬

Features • Installation • Quick Start • Contributing

🔥 Features

• ✅ Fully‑typed (TypeScript)
• 📡 Stream image moderation
• 🌐 URL image inputs
• ⚙️ Threshold presets + custom thresholds
• 🛠️ Helper utilities: isNSFW(), listFlaggedCategories()

📦 Installation

npm install sightengine-js
# or
yarn add sightengine-js

🚀 Quick Start

import { SightEngineClient, isNSFW, listFlaggedCategories } from "sightengine-js";

const sightengine = new SightEngineClient({
	apiSecret: "<your api secret>",
	apiUser: "<your api user>",
});

const imageModerationResult = await sightengine.moderateImage(
	"<path to your image>",
	[
		"nudity-2.1",
		"gore-2.0",
		"offensive-2.0",
	],
);

// URL-based
const urlImageModerationResult = await sightengine.moderateImageByUrl(
	"<image URL>",
	[
		"nudity-2.1",
		"gore-2.0",
		"offensive-2.0",
	],
);

console.log(listFlaggedCategories(imageModerationResult));
console.log(isNSFW(imageModerationResult));

📚 API Reference

new SightEngineClient(options)

Property	Type	Description
apiUser	string	Your SightEngine API user.
apiSecret	string	Your SightEngine API secret.

client.moderateImage(path, models?)

• path – Local file system path to the image.
• models – Array of model identifiers (default: ["nudity-2.1"]).

Returns a promise that resolves to ImageModerationResponse.

client.moderateImageByUrl(url, models?)

• url – Direct URL to the image.
• models – Same as above.

Returns a promise that resolves to ImageModerationResponse.

Types

• ImageModerationResponse – Raw response from SightEngine (see src/types.ts).
• NormalizedImageModerationResponse – Internal normalized shape used by helpers.

⚙️ Thresholds & Presets

import { DEFAULT_THRESHOLDS, PRESET_THRESHOLDS } from "sightengine-js";

// Example: use the strict preset
const thresholds = PRESET_THRESHOLDS.strict;

Model	Default
nudity	0.85
offensive	0.5
gore	0.5
violence	0.5
self-harm	0.5

🛠️ Helper Utilities

• isNSFW(response, threshold?) – Returns true if any of the important categories (nudity, gore, violence, self-harm, sexual) exceed the given threshold.
• listFlaggedCategories(response, thresholds?) – Returns an array of categories (and sub‑scores) that exceed the supplied thresholds.

Both functions accept the raw ImageModerationResponse from the client.

🤝 Contributing

1. Fork the repo
2. Create a feature branch (git checkout -b feat/...)
3. Commit with Commitizen (npm run commit)
4. Push & open a PR against develop
5. Tests must pass, coverage ≥ 90%

---

📜 License

MIT © Ali Nazari

---

_{Built with ❤️ by Ali Nazari, for developers. Happy encoding! 🎬}

⭐ Star us on GitHub • 🐦 Follow on Linkedin