# AI Content

## Overview

| Property | Value |
|----------|-------|
| **Status** | private |
| **Method** | `GET` |
| **Endpoint** | `/v1/analyze` |
| **Base URL** | `https://api.capturekit.dev` |
| **Credit Cost** | 1 credit per call |
| **Documentation** | https://www.piloterr.com/library/analyze |

## Description

AI Content API analyzes any webpage and returns AI insights summary, value proposition, features, categories, use cases, and optional contact signals.

## Authentication

- **Key Name:** `x-api-key`
- **Location:** HTTP Header
- **Get an API key:** https://app.capturekit.dev/register

## Example Request

```bash
curl --location --request GET 'https://api.capturekit.dev/v1/analyze' \
  --header 'Content-Type: application/json' \
  --header 'x-api-key: YOUR_API_KEY'
```

## Example Response

### Example

```json
{
  "data": {
    "emails": [
      "admin@company.com"
    ],
    "summary": "Tailwind CSS is a utility-first CSS framework for modern web design.",
    "useCases": [
      "Building websites",
      "Customizing themes",
      "Managing CSS specificity"
    ],
    "categories": [
      "Web Development",
      "CSS Frameworks",
      "Frontend Development",
      "Responsive Design"
    ],
    "keyFeatures": [
      "Responsive design",
      "Dark mode",
      "Animations",
      "Grid layout"
    ],
    "phoneNumbers": [
      "+13322767084"
    ],
    "targetAudience": "Web developers and designers.",
    "valueProposition": "A flexible CSS framework packed with utility classes."
  },
  "success": true
}
```

## Documentation

## Overview

CaptureKit’s **AI Content API** analyzes a webpage and returns **AI-generated insights** such as a **summary**, **value proposition**, **key features**, **categories**, **use cases**, and optionally **contact signals** (emails, phone numbers).

It’s built for workflows where you need fast, consistent **content intelligence** for **competitive research**, **lead enrichment**, **content monitoring**, and **automation**, without maintaining your own scraping + LLM pipeline.

**API endpoint:** `https://api.capturekit.dev/v1/analyze`

## Typical use cases

- **Competitive research:** summarize competitor landing pages and extract positioning/value propositions.
- **Product intelligence:** extract key features, categories, and common use cases from product pages.
- **Lead enrichment:** identify target audience and extract public emails / phone numbers when present.
- **Content monitoring:** track changes in messaging, features, and positioning over time.
- **SEO & content ops:** quickly classify pages and generate consistent summaries at scale.
- **Custom AI analysis:** inject your own fields and instructions via `custom_response` and `custom_prompt`.
- **Geo / restricted pages:** route requests through a proxy.

## Quickstart

Send a request with `url` and your `x-api-key`.

Example request:

```bash
curl -L "https://api.capturekit.dev/v1/analyze?x-api-key=YOUR_KEY&url=https://tailwindcss.com"
```

## Parameters

### Required

- **url** *(string, required)*  
  The URL of the webpage to analyze.

- **x-api-key** *(string, required)*  
  Your API access key. Can be provided via the `x-api-key` query parameter, `x-api-key` header, or request body.

### AI output customization

- **custom_response** *(object | string, optional)*  
  Add custom fields to be included in the AI response. You can provide it:
  - as a JSON object in the request body, or
  - as a JSON-stringified object in query parameters.

  Example:
  ```json
  {
    "sentiment": "sentiment",
    "language": "language"
  }
  ```

- **custom_prompt** *(string, optional)*  
  Add your own instructions to guide the AI’s analysis (tone, focus, constraints, etc.).

### Routing / caching

- **proxy** *(string, optional)*  
  Specify a proxy server to route your request through. Supports HTTP, HTTPS, and SOCKS5 proxies.  
  Format: `http://username:password@proxy.com:PORT`.

- **cache** *(boolean, optional, default: `false`)*  
  Cache the response.

- **cache_ttl** *(number, optional, default: `2592000`)*  
  Cache the response for a custom TTL (in seconds).  
  Maximum `2592000` seconds (1 month), minimum `3600` seconds (1 hour).

## Notes / limits

- **AI outputs are probabilistic:** results can vary slightly over time; store timestamps and consider caching for monitoring workflows.
- **Custom fields:** keep `custom_response` focused (few fields) for predictable outputs.
- **Proxy usage:** use `proxy` for geo-restricted pages or when rotating IPs is required.
- **Caching:** enable `cache` + `cache_ttl` for repeated audits/monitoring to reduce cost and latency.