Pornhub MCP Server

Instructions

Implementation Reference

• server.py:385-416 (handler)

  Main implementation of the historical_analysis tool. Decorated with @mcp.tool(), it analyzes historical trends for a keyword across multiple years using Google Trends data. Handles input validation (year >= 2004, start_year < end_year) and formats the timeframe before delegating to search_trends helper.
  @mcp.tool() async def historical_analysis( keyword: str, start_year: int = 2020, end_year: int = 2024, region: str = "US" ) -> str: """ Analyze historical trends for a keyword across multiple years. Args: keyword: Single keyword to analyze (performer, platform, or category) Example: "Lana Rhoades", "onlyfans", "milf" start_year: Starting year (2004 or later) end_year: Ending year (default: 2024) region: Region code (default: US) Returns: Multi-year trend analysis showing growth, peaks, and patterns. """ if start_year < 2004: return "⚠️ Google Trends data is only available from 2004 onwards." if start_year >= end_year: return "⚠️ Start year must be before end year." timeframe = f"{start_year}-01-01 {end_year}-12-31" result = await search_trends([keyword], timeframe, region) return f"📅 Historical Analysis: {start_year}-{end_year}\n\n{result}"
• server.py:245-299 (helper)

  Core search_trends helper function called by historical_analysis. Fetches and formats Google Trends data for given keywords, timeframe, and region. Validates inputs (max 5 keywords), calls get_trends_data, and formats results with interest over time, regional data, and related queries.
  async def search_trends( keywords: list[str], timeframe: str = "today 12-m", region: str = "US" ) -> str: """ Search Google Trends for any keywords (performers, platforms, categories, etc.). Args: keywords: List of search terms to analyze (max 5). Examples: ["Lana Rhoades", "Riley Reid"] ["pornhub", "onlyfans"] ["milf", "teen", "amateur"] timeframe: Time period. Options: 'today 12-m' (past year, default) 'today 3-m' (past 3 months) 'today 5-y' (past 5 years) '2020-01-01 2024-12-31' (custom date range) Available back to 2004 region: Geographic region code: 'US' (USA, default) 'GB' (UK) '' (Worldwide) Any ISO country code Returns: Complete Google Trends analysis with interest over time, regional data, and related queries. """ if not pytrends: return "❌ Google Trends API is not available. Please check configuration." if len(keywords) > 5: return "⚠️ Maximum 5 keywords allowed per query. Please reduce your list." if len(keywords) == 0: return "⚠️ Please provide at least one keyword to search." # Fetch data data = get_trends_data(keywords, timeframe, region) # Format and return results result = [ format_interest_over_time(data), "", format_regional_interest(data, top_n=10), format_related_queries(data), "", "📝 Notes:", "- Values are on a 0-100 scale where 100 = peak popularity for the time period", "- Data represents search interest, not absolute search volumes", f"- Data fetched: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}" ] return "\n".join(result)
• server.py:42-109 (helper)

  Low-level get_trends_data helper function that interfaces with the Google Trends API (pytrends). Builds API payload, fetches interest over time, regional data, and related queries. Implements caching (1 hour) and rate limiting to avoid excessive API calls.
  def get_trends_data(keywords: list[str], timeframe: str = 'today 12-m', geo: str = 'US') -> dict: """ Fetch Google Trends data for given keywords. Args: keywords: List of search terms to compare (max 5) timeframe: Time period (e.g., 'today 12-m', 'today 5-y', '2020-01-01 2024-12-31') geo: Geographic region (e.g., 'US', 'GB', '' for worldwide) Returns: Dictionary with trends data """ if not pytrends: return {"error": "Google Trends API not available"} # Check cache cache_key = f"{','.join(keywords)}_{timeframe}_{geo}" if cache_key in TRENDS_CACHE: cached = TRENDS_CACHE[cache_key] age = (datetime.now() - datetime.fromisoformat(cached['fetched_at'])).seconds if age < 3600: # Cache for 1 hour print(f"Using cached data (age: {age}s)", file=sys.stderr) return cached try: print(f"Fetching Google Trends: {keywords}, {timeframe}, {geo}", file=sys.stderr) # Build payload pytrends.build_payload(keywords, cat=0, timeframe=timeframe, geo=geo, gprop='') # Get interest over time interest_over_time_df = pytrends.interest_over_time() # Get interest by region try: interest_by_region_df = pytrends.interest_by_region(resolution='REGION', inc_low_vol=True, inc_geo_code=False) except Exception as e: print(f"Could not fetch regional data: {e}", file=sys.stderr) interest_by_region_df = pd.DataFrame() # Get related queries try: related_queries = pytrends.related_queries() except Exception as e: print(f"Could not fetch related queries: {e}", file=sys.stderr) related_queries = {} result = { "keywords": keywords, "timeframe": timeframe, "geo": geo, "interest_over_time": interest_over_time_df.to_dict() if not interest_over_time_df.empty else {}, "interest_by_region": interest_by_region_df.to_dict() if not interest_by_region_df.empty else {}, "related_queries": related_queries, "fetched_at": datetime.now().isoformat() } # Cache the result TRENDS_CACHE[cache_key] = result # Rate limiting time.sleep(1) return result except Exception as e: print(f"Error fetching Google Trends data: {e}", file=sys.stderr) return {"error": str(e)}