# Check AI chatbot replies for toxicity before posting

**Use case:** 

Use this task as a guardrail for AI agents and chatbots. Send the replies that your bot will post, and get toxicity, insult, threat and identity attack scores from 0 to 1. The task uses a strict threshold of 0.3, so it flags borderline replies too. Hold flagged replies for a manual check. Fast, deterministic and no LLM. English only.

## Input

```json
{
  "texts": [
    "Thanks for reaching out! Your refund was approved and will arrive in 3 to 5 business days.",
    "I understand the frustration. Let me check the order status for you.",
    "Honestly, only an idiot would ask a question this stupid.",
    "Sure, here is a summary of the article in three bullet points.",
    "Your complaint is worthless and so are you. Stop wasting my time."
  ],
  "attributes": [
    "TOXICITY",
    "INSULT",
    "THREAT",
    "IDENTITY_ATTACK"
  ],
  "threshold": "0.3"
}
```

## Output

```json
{
  "text": {
    "label": "Text",
    "format": "string"
  },
  "flagged": {
    "label": "Flagged",
    "format": "boolean"
  },
  "scores.TOXICITY": {
    "label": "Toxicity",
    "format": "number"
  },
  "scores.SEVERE_TOXICITY": {
    "label": "Severe toxicity",
    "format": "number"
  },
  "scores.INSULT": {
    "label": "Insult",
    "format": "number"
  },
  "scores.PROFANITY": {
    "label": "Profanity",
    "format": "number"
  },
  "scores.THREAT": {
    "label": "Threat",
    "format": "number"
  },
  "scores.IDENTITY_ATTACK": {
    "label": "Identity attack",
    "format": "number"
  }
}
```

## About this Actor

This example demonstrates how to use [Toxic Comment Detector - Content Moderation API](https://apify.com/koyourmoon/toxicity-detector.md) with a specific input configuration. Visit the [Actor detail page](https://apify.com/koyourmoon/toxicity-detector.md) to learn more, explore other use cases, and run it yourself.


## How to integrate an Actor?

This Task's input is already configured above. Use it as-is rather than inventing a new one.

Do not guess an integration path. Every one of them is in the agent quickstart at https://apify.com/agents.md: the Apify MCP server, Agent Skills with the Apify CLI, the JavaScript and Python clients, the REST API, and the account-free path for an agent with no human to sign in. It also carries the rule on stating cost before the first paid run.

For full API examples (JavaScript, Python, CLI, MCP, OpenAPI), see this Task's Actor page: https://apify.com/koyourmoon/toxicity-detector.md

Each client library has reference documentation the quickstart does not restate: [JavaScript/TypeScript](https://docs.apify.com/api/client/js/docs.md) (`npm install apify-client`) and [Python](https://docs.apify.com/api/client/python/docs.md) (`pip install apify-client`).