Exposes fetch and search tools that convert web pages to clean markdown using a 15-strategy fallback pipeline. Handles site-specific formats like tweets, YouTube transcripts, arXiv papers, PDFs, and GitHub READMEs before falling back to Jina Reader, archive services, and stealth fetching. Routes requests through agentsweb.org's shared cache first for sub-50ms hits on previously fetched URLs. No API keys required for basic fetching, though you'll want Brave or SearXNG for reliable search and Cloudflare Browser Run for JavaScript-heavy sites. Includes research-topic and extract-article prompts that chain search with multi-source fetching. Reach for this when you need to give Claude reliable web access without hitting 403s or wrestling with raw HTML.
claude mcp add --transport stdio bighippoman-intercept-mcp uvx intercept-mcp