Parallel Web

A unified web toolkit that routes between search, URL extraction, data enrichment, and deep research, all powered by parallel-cli. The standout feature is its academic source priority: it defaults to peer-reviewed papers, preprints, and scholarly databases over general web results when the topic warrants it. Use web search for quick lookups, extract for fetching specific URLs (including academic PDFs), enrichment when you have a list of entities that need the same information added, and deep research only when someone explicitly asks for exhaustive reports. The routing logic is sensible: it knows not to loop web searches when enrichment would handle batch lookups better. Requires parallel-cli installed and authenticated, but the setup instructions are straightforward.

Install to Claude Code

npx -y skills add k-dense-ai/scientific-agent-skills --skill parallel-web --agent claude-code