v3.0.0

[3.0.0] - 2026-10-11

Removed

• 46 web_data_* tools (Amazon, LinkedIn, Instagram, TikTok, YouTube,
  npm, PyPI and others). The scraper tools cover the same scrapers and
  1,200+ more: search_scrapers → get_scraper_details →
  run_scraper.
• web_data_grok_ai_insights and web_data_reuter_news. Their
  scrapers no longer exist, so these tools always failed.

Changed

• web_data_chatgpt_ai_insights and web_data_perplexity_ai_insights
  are renamed to chatgpt_ai_insights and perplexity_ai_insights.
• Existing configs keep working. Old group IDs (ecommerce, social,
  business, finance, research, app_stores, travel, code)
  now enable the scraper tools, and old web_data_* names in TOOLS
  enable the scraper tools or the renamed AI tools.
• Scraper tool requests send the scraper's dataset ID and method in
  x-mcp-dataset-id and x-mcp-method headers, so failed runs can be
  traced to a scraper.
• If you installed the bright-data-mcp, competitive-intel or
  price-comparison skills, update them: older versions tell the agent
  to use the removed tools.