Skill Builder

It auto-detects whether you're pointing it at docs, a GitHub repo, PDFs, videos, or local code, then runs the right scraper from its 40-tool arsenal. The workflow is sensible: generate a config, estimate scope, scrape, optionally enhance with AI, then package or export to vector databases like Weaviate or Qdrant. What's nice is it handles the tedious parts of documentation conversion without making you think too hard about formats. If you're building context packages for LLMs or need to sync skills with upstream docs, this does the grunt work.