この紹介文は現在英語版のみです。
Overview
Deepcrawl is an agent-oriented platform for extracting website context. It returns the cleaned Markdown of a page’s content, an agent-friendly hierarchical links tree and metadata that LLMs can digest with minimal tokens — reducing context switching and hallucination.
It is positioned as a free, open-source alternative to Firecrawl. The whole platform is open: a Next.js dashboard, API Workers, auth Workers and the database.
Scope
Deepcrawl is built for high-frequency agent workloads that read public pages. Its README states that it does not target anti-scraping or anti-bot use.
Status
The project is under rapid development, and its authors warn against using it in production yet. Documentation lives at deepcrawl.dev/docs.