scrapecreators-mcp-server

ScraperCreators MCP Server

A FastMCP server for the ScraperCreators platform that provides MCP tools and resources for content scraping operations.

Project Structure

• src/mcp_server/ - Core server implementation

  • main.py - FastMCP server with tools
  • cli.py - CLI utility for version information

• src/scrapecreator_api/ - ScraperCreators API integration

  • reddit_search.py - Reddit Search API client
  • response_models.py - Response data models

• test_server.py - Direct integration tests

• run_server.sh - Script to start the server

Installation

# Create a virtual environment (optional but recommended)
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install the package and dependencies
pip install -e .

Development

# Install development dependencies
pip install -e ".[dev]"

# Run tests
pytest

Usage

Running the server

The project provides a FastMCP server that can be run using:

./run_server.sh

This uses the FastMCP CLI to run the server with the stdio transport.

Alternatively, you can run it directly with:

# Using Python
python -m src.mcp_server.main

# Using FastMCP CLI
fastmcp run src/mcp_server/main.py:mcp

Reddit Search API

The RedditSearch class provides a Python client for the ScrapeCreators Reddit Search API.

Basic Usage

from scrapecreator_api.reddit_search import RedditSearch

# Initialize with API key (or set REDDIT_API_KEY in .env.local)
client = RedditSearch(api_key="your_api_key")

# Basic search
results = client.search(query="python programming")

# Search with modifiers
results = client.search(
    query="fastapi tutorial",
    sort="new",
    timeframe="week",
    subreddit="learnpython",
    self=True
)

# Search with pagination and limit
results = client.search_with_pagination(
    query="data science",
    limit=50
)

Search Parameters

• query: Search keywords or phrases
• sort: Sort method (relevance, new, top, comment_count)
• timeframe: Time period (all, day, week, month, year)
• after: Pagination token from previous response
• Modifiers:
  • author: Filter by author
  • subreddit: Filter by subreddit
  • title: Search in titles
  • selftext: Search in post content
  • flair: Filter by post flair
  • url: Search in URLs
  • self: Filter self posts (boolean)

Search Response Options

The API supports two response modes:

Inline Mode

Returns JSON response directly with optional result limiting:

response = client.search(
    query="your query",
    return_mode="inline",
    max_results=10  # Optional limit
)

File Mode

Saves results to file and returns metadata:

response = client.search(
    query="your query",
    return_mode="file",
    output_dir="/custom/path"  # Optional custom directory
)

Both modes return a SearchResponse object containing:

• success: Boolean indicating success
• count: Number of posts found
• posts: List of RedditPost objects (inline mode only)
• file_path: Path to saved JSON file (file mode only)

Error Handling

The client provides custom exceptions for different error scenarios:

• RedditSearchError: Base exception class
• RedditSearchConnectionError: Connection issues
• RedditSearchAuthenticationError: Invalid API key
• RedditSearchAPIError: API response errors

Testing

The recommended way to test the server is through direct integration:

./test_client.sh

This runs a test that imports the MCP instance directly and tests the tools without requiring a separate server process. Based on the FastMCP quickstart guide, this is the most reliable approach for testing.

Available Tools

• hello: Returns a greeting message with the given name

Notes

• The server instance is named mcp in main.py to make it easily discoverable by the FastMCP CLI
• For production use, consider using HTTP or other transports instead of stdio
• API keys should be stored in .env.local file or provided as environment variables