npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2024 – Pkg Stats / Ryan Hefner

content-checker

v1.1.3

Published

Open-source and modern AI content moderation

Downloads

267

Readme

Content-Checker (by OpenModerator)

Release Content Checker npm

content-checker is designed to be a modern, open-source library for programmatic and AI content moderation. Currently content-checker supports image and text moderation. Thanks to LLMs in addition to detecting specific profane words, we can detect malicious intent in text. So, a user who tries to circumvent the AI profanity filter by using a variation of a profane word, or even just a malicious phrase without a specific word in the profanity list, will still be flagged. Image moderation is also supported, using the Inception V3 model of the NSFW JS library.

Future features will include moderation tools (auto-ban, bots), more powerful models, and multimedia support for video and audio moderation.

How It Works

content-checker builds on the popular bad-words package, but updated to use TypeScript and ES6, and merging in badwords-list, which it used as a dependency.

The AI moderation is powered by a database of profane words, multiple LLMs for text analysis, and the NSFW JS library for image analysis. The models will likely be changed in the future as more powerful models become available. The importance of AI moderation for text is we need to be able to detect malicious intent, not just specific words.

The API will return whether text is profane or not, but note that different types of profanity can be detected, and the exact types returned will depend on the model used. Currently, the hosted model uses around a 60% confidence threshold for a profanity detection when using Google's Perspective API or Google's Natural Language API.

Unsafe text:

{
  "profane": true,
  "type": ["TOXICITY", "SEVERE_TOXICITY"]
}

Safe text:

{
  "profane": false,
  "type": []
}

The API returns responses in the following format for image moderation (the possible image types for now include "Porn" and "Hentai"). If the highest probability category is a NSFW category like the ones mentioned then the image will be flagged as NSFW:

Unsafe image:

{
  "nsfw": true,
  "type": ["Porn"]
}

Safe image:

{
  "nsfw": false,
  "type": []
}

Note that type is an array, so it can contain multiple types of profanity or none at all (if it hits the fine-tuned model). The API is rate limited to 10 requests from the same IP address per 10 seconds. If you need more, please contact me.

Installation

Use npm to install content-checker.

npm install content-checker

Table of Contents

  1. Standard Text Moderation
  2. AI Text Moderation
  3. AI Image Moderation

Usage

Standard Text Moderation

Initialize a filter

import { Filter } from "content-checker";
// Or for CommonJS: const { Filter } = require('content-checker');
const filter = new Filter();

console.log(filter.clean("Don't be an ash0le")); //Don't be an ******

Placeholder Overrides

import { Filter } from "content-checker";
const customFilter = new Filter({ placeHolder: "x" });

customFilter.clean("Don't be an ash0le");

Regex Overrides

const filter = new Filter({ regex: /\*|\.|$/gi });

const filter = new Filter({ replaceRegex: /[A-Za-z0-9가-힣_]/g });

Add words to the blacklist

const filter = new Filter();

filter.addWords("some", "bad", "word");

filter.clean("some bad word!");

// or use an array using the spread operator

const newBadWords = ["some", "bad", "word"];

filter.addWords(...newBadWords);

filter.clean("some bad word!"); //**** *** ****!

// or

const filter = new Filter({ list: ["some", "bad", "word"] });

filter.clean("some bad word!"); // **** *** ****!

Instantiate with an empty list

const filter = new Filter({ emptyList: true });

filter.clean("hell this wont clean anything"); // hell this wont clean anything

Remove words from the blacklist

const filter = new Filter();

filter.removeWords("hells", "sadist");

filter.clean("some hells word!"); // some hells word!

// or use an array using the spread operator

let removeWords = ["hells", "sadist"];

filter.removeWords(...removeWords);

filter.clean("some sadist hells word!");

AI Text Moderation

To use AI text moderation, ensure you have the OPEN_MODERATOR_API_KEY set in your environment variables (one can be generated for free at www.openmoderator.com) or passed as a parameter during the initialization of the Filter class.

Initialize a filter

import { Filter } from "content-checker";

const filter = new Filter({ openModeratorAPIKey: "your_api_key_here" });

Check a string for profanity

The isProfaneAI method checks if a string contains profane language using AI. It returns a promise that resolves to an object containing a profane boolean and a type array which lists the categories of profanity detected (e.g., "harassment").

const config = {
  // checkManualProfanityList is optional and defaults to false; it checks for the words in lang.ts (if under 50 words) before hitting the AI model. Note that this affects performance.
  checkManualProfanityList: false,
  // provider defaults to "google-perspective-api" (Google's Perspective API); it can also be "openai" (OpenAI Moderation API) or "google-natural-language-api" (Google's Natural Language API)
  provider: "google-perspective-api",
};

filter.isProfaneAI("your string here", config).then((response) => {
  if (response.profane) {
    console.log("Profanity found. Types: ", response.type.join(", "));
  } else {
    console.log("No profanity found");
  }
});

AI Image Moderation

To use AI image moderation, ensure you have the OPEN_MODERATOR_API_KEY set in your environment variables (one can be generated for free at www.openmoderator.com) or passed as a parameter during the initialization of the Filter class. For now the NSFW JS library is used for image moderation, but this will be replaced with a more powerful model in the future. Ensure you're uploading either a PNG or JPEG image.

Check an image for NSFW content

Raw JS example:

const imageInput = document.getElementById("imageInput");

imageInput.addEventListener("change", async (event) => {
  const file = event.target.files[0];
  if (file) {
    try {
      const response = await filter.isImageNSFW(file);
      if (response.nsfw) {
        console.log("NSFW content detected. Types:", response.type.join(", "));
      } else {
        console.log("Image is safe.");
      }
    } catch (error) {
      console.error("Error checking image:", error);
    }
  }
});

React example:

import React, { useState } from "react";
import { Filter } from "content-checker";

const ImageModeration = () => {
  const [image, setImage] = useState(null);
  const [moderationResult, setModerationResult] = useState("");

  const filter = new Filter({ openModeratorAPIKey: "your_api_key_here" });

  const handleImageChange = (event) => {
    setImage(event.target.files[0]);
  };

  const checkImage = async () => {
    if (image) {
      try {
        const response = await filter.isImageNSFW(image);
        if (response.nsfw) {
          setModerationResult(
            `NSFW content detected. Types: ${response.type.join(", ")}`,
          );
        } else {
          setModerationResult("Image is safe.");
        }
      } catch (error) {
        console.error("Error checking image:", error);
        setModerationResult("Error occurred while checking the image.");
      }
    }
  };

  return (
    <div>
      <input
        type="file"
        onChange={handleImageChange}
        accept="image/png, image/jpeg"
      />
      <button onClick={checkImage}>Check Image</button>
      {moderationResult && <p>{moderationResult}</p>}
    </div>
  );
};

export default ImageModeration;

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

Please make sure to update tests as appropriate.

Authors

  • Jacob Habib (@jahabeebs), founder of OpenModerator