Research Skills

Web search, deep research, and content extraction for OpenClaw

Research skills help OpenClaw search the web, run deep research, summarize content, and fetch protected pages. Install from ClawHub with openclaw skills install <skill-name>.

Web Search and Research

  • Deep Research Agent - Deep research and web search. Research use case
  • Web search - Web search and browsing
  • Summarizer - Text summarization

Content Extraction and Browser

  • GhostFetch - Stealth web fetch for X.com, LinkedIn, protected sites; bypasses 403/Cloudflare, returns Markdown. GhostFetch guide
  • PinchTab - Standalone HTTP browser bridge. Token-efficient, multi-instance. PinchTab guide · Browser guide
  • camofox-browser - Headless browser plugin for X, Product Hunt, Amazon. camofox guide

Install and enable

  1. Search ClawHub for skills in this category.
  2. Follow safe install checklist before enabling untrusted packages.
  3. openclaw skills install <skill-name> then openclaw skills list.
  4. Restart the Gateway if the skill README requires it.

Other Categories

Productivity - Calendar, Gmail, Todoist. Dev and Infrastructure - GitHub, Docker, n8n. Media - Spotify, Sonos, voice. Channels - Slack, Discord, WhatsApp. All Skills - Full list.

Provider keys for research skills

Many research skills call external search or fetch APIs (Brave, Firecrawl, etc.). Map keys in Configuration and prefer read-only scopes. For protected sites, combine GhostFetch or PinchTab with protected sites guidance instead of giving the agent your personal browser session.

Cost control

Research skills trigger search and fetch tools—expensive in loops. Disable deep-research skills in public channels; run them from paired DMs or scheduled cron with a spend-capped model.

See also