The AI Scientist's Supply Cabinet

What it does

This is a curated awesome-list that catalogs AI tools, libraries, datasets, and frameworks pitched at scientific discovery. It sorts resources across the entire research lifecycle—literature search, document parsing, data visualization, paper-to-code reproduction, presentation generation, and domain-specific modeling in biology, chemistry, physics, and climate science. Think of it as a supply cabinet for researchers who would rather automate the grunt work than drown in it.

The interesting bit

The list treats science as an end-to-end pipeline to be optimized. It shelves niche utilities like Paper2Poster and Paper2Video beside industrial-strength PDF parsers such as MinerU and Docling, acknowledging that modern research toil now spans everything from citation graphs to PowerPoint animations.

Key highlights

• Covers the full research automation chain: literature search, PDF parsing, chart generation, code extraction, poster and slide creation, plus autonomous research agents.
• Heavy focus on document intelligence, listing Nougat, olmOCR, PDF-Extract-Kit, and IBM’s Docling for converting scientific PDFs into structured, LLM-ready formats.
• Includes agent skill packs such as Google DeepMind’s science-skills and K-Dense-AI’s scientific-agent-skills meant to plug into coding assistants.
• Domain-specific sections span biology, chemistry, materials, physics, astronomy, earth science, and agriculture.
• MIT licensed and offered in nine languages.

Caveats

• The source README is truncated, suggesting the list is extensive enough to feel unwieldy.
• Some utilities like Chat2Plot and AutoViz appear under multiple headings, so the category boundaries are not always crisp.

Verdict

Researchers who spend more time formatting slides than running experiments will find this a useful compass. If you need a single integrated platform rather than a curated directory, look elsewhere.