ContextFetch
Perform web searches, extract web pages as LLM-ready Markdown, and run automated research workflows for AI agents.
ContextFetch equips AI agents with cost-effective web search and content extraction engineered specifically for LLM context windows. By stripping web clutter and pruning pages down to cross-encoder-scored, relevant Markdown passages, it delivers high-density factual context without token waste. The server includes 11 specialized tools for standard search, URL extraction, multi-query server-side research, site crawling, and factual question answering, accompanied by a keyless free tier for immediate integration.
주요 기능
01Server-side multi-query research compiling deduplicated, ranked evidence packs
02Direct URL extraction optimized for public articles, technical docs, and code files
03Token-efficient web search with cross-encoder passage pruning into clean Markdown
04Targeted web crawling and sitemap discovery with customizable path regexes
052 GitHub stars
06Zero-configuration keyless free tier offering 25 daily searches with no signup
사용 사례
01Crawling documentation hubs and technical blogs to feed clean Markdown into RAG pipelines
02Equipping coding assistants like Claude Code and Cursor with real-time documentation search
03Conducting automated multi-angle research to synthesize factual evidence packs for reports