Skip to main content
Kilo is the all-in-one agentic engineering platform with support for MCP (Model Context Protocol) servers. The Olostep MCP Server integrates seamlessly with Kilo, giving your AI agent powerful web data capabilities.

Features

The Olostep MCP Server provides access to 5 core Olostep capabilities:

Scrape Website

Extract content from any single URL in multiple formats (Markdown, HTML, JSON, text)

Search Web

Search the web and get structured, parser-based results

Answers (AI)

Get AI-powered answers with natural language queries and citations

Batch Scrape URLs

Process up to 10,000 URLs in parallel. Perfect for large-scale data extraction

Crawl Websites

Autonomously discover and scrape entire websites by following links

Installation

1. Get Your API Key

Get your Olostep API key from the Olostep Dashboard.

2. Configure Kilo

In your project directory, create or edit .kilocode/mcp.json to add the Olostep MCP Server:
{
  "mcpServers": {
    "olostep": {
      "url": "https://mcp.olostep.com/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_OLOSTEP_API_KEY"
      }
    }
  }
}
Replace YOUR_OLOSTEP_API_KEY with your actual API key from the Olostep Dashboard.

3. Verify Installation

Run Kilo and the Olostep tools should be automatically available:
kilo

Available Tools

Once configured, the following Olostep tools are available to your Kilo agent:

scrape_website

Extract content from a single URL. Supports multiple formats and JavaScript rendering. Example prompt:
Scrape the content from https://example.com and give me the markdown version
Parameters:
  • url - Website URL to scrape (required)
  • format - Output format: markdown, html, json, or text (default: markdown)
  • country - Country code for location-specific content (e.g., “US”, “GB”)
  • wait_before_scraping - Wait time in milliseconds for JavaScript rendering (0-10000)
  • parser - Optional parser ID for specialized extraction (e.g., “@olostep/amazon-product”)

search_web

Search the web and return structured, parser-based results. Example prompt:
Search for the latest news about AI regulations in the EU
Parameters:
  • query - Search query (required)
  • country - Country code for location-specific results

answers

Get AI-powered answers from web searches with sources and citations. Example prompt:
What are the top 5 emerging AI trends in 2026? Get me structured answers with sources.
Parameters:
  • query - Natural language question (required)
  • format - Output format for answers (JSON structure of your choice)
  • country - Country code for location-specific searches

batch_scrape_urls

Process multiple URLs in parallel. Perfect for large-scale data extraction. Example prompt:
Scrape these 50 product URLs and extract title, price, and description in JSON format
Parameters:
  • urls - Array of URLs to scrape (required)
  • format - Output format for all URLs
  • country - Country code for location-specific content
  • parser - Optional parser ID for all URLs

create_crawl

Autonomously discover and scrape entire websites by following links from a start URL. Example prompt:
Crawl the website starting from https://example.com and extract all product pages
Parameters:
  • start_url - Starting URL for the crawl (required)
  • max_pages - Maximum number of pages to crawl (default: 10)
  • follow_links - Whether to follow links (default: true)

create_map

Extract all URLs from a website for site structure analysis and content discovery. Example prompt:
Map all URLs on https://example.com to understand the site structure
Parameters:
  • website_url - Website to map (required)
  • search_query - Optional query to filter results
  • top_n - Limit number of returned URLs

Example Workflows

Let Kilo agent gather information from the web:
Use the search_web tool to find recent articles about LLM safety,
then use the answers tool to synthesize the key findings with sources.
Give me a comprehensive summary.

Product Data Extraction

Scrape multiple product pages:
I have 100 product URLs from an e-commerce site. Use batch_scrape_urls
to extract title, price, description, and availability from each one.
Format the results as JSON.

Website Discovery and Analysis

Crawl and analyze a website structure:
I want to understand the structure of https://example.com.
First, use create_map to get all URLs, then crawl the main sections
and give me a summary of the site's content organization.

Environment Variables

When using the local/stdio configuration, make sure your environment has the API key set:
export OLOSTEP_API_KEY="your_api_key_here"
kilo
Or in your .env file:
OLOSTEP_API_KEY=your_api_key_here

Troubleshooting

Tools not appearing in Kilo:
  • Verify the .kilocode/mcp.json is properly formatted (valid JSON)
  • Ensure your API key is correct in the configuration
  • Try restarting Kilo after configuration changes
API key authentication errors:
  • Double-check your API key from the Olostep Dashboard
  • Make sure there are no extra spaces or special characters in the key
Remote hosted MCP not connecting:
  • Verify your internet connection
  • Check that https://mcp.olostep.com/mcp is accessible
  • Ensure the Authorization header format is correct: Bearer YOUR_KEY

Learn More