Skip to main content

​
Overview

WraithBytes maintains 50+ custom web parsers that automatically detect the platform from your URL and apply specialized extraction logic. Parsers are built through statistical analysis of real page data — as requests accumulate, the system identifies consistent patterns, filters out noise, and uses AI to generate optimized parsers unique to each site. These parsers are continuously monitored and auto-regenerated when sites change. AI mode can be forced on any request for instant intelligent extraction.

​
How Auto-Detection Works

When you call POST /api/v1/internet/fetch/, the URL host is analyzed to select the appropriate parser:
URL PatternParser
linkedin.com/in/...LinkedIn Profile
linkedin.com/company/...LinkedIn Company
linkedin.com/jobs/...LinkedIn Jobs
amazon.com/dp/...Amazon Product
instagram.com/...Instagram
Any other URLGeneral-Purpose
All parsers support both structured JSON and Markdown output — use the markdown or json fetch options to control the format. No configuration needed — just send the URL.

​
Parser Categories

LinkedIn

Profiles, companies, and job listings with rich structured output

Amazon

Product pages and customer reviews

Instagram

Profiles and content

General Web

Any URL — Markdown, links, images, or custom JSON schema extraction

​
Response Structure

All parsers return responses in the same envelope: 
{
  "success": true,
  "status": "success",
  "data": {
    // Parser-specific fields
  }
}
The data object structure varies by parser. See individual parser pages for field definitions.

​
Data Characteristics

​
Encoding

  • UTF-8 for all text fields
  • Supports international characters

​
Timestamps

  • Format: ISO 8601 (YYYY-MM-DDTHH:mm:ssZ)
  • Timezone: UTC

​
Nullability

  • Most fields can be null if not present on the source page
  • Always present: core identifiers and timestamps