All skills
nature-literature-pipeline
Run a daily literature pipeline that searches, scores, reads, delivers, and archives research papers.
- Part of Yuan1z0825/nature-skills
- 13.2k installs
- Learning Resources · Study tools
When to use it
You already have research keywords and a place to send digests, but manually scanning arXiv and OpenAlex each day is eating your time.
For example
You say
Use nature-literature-pipeline to set up a daily 08:30 push on protein language models, 30 candidates, top 5 to my Feishu group, archive to ~/vault/raw.
You get
- Configured keyword and scoring setup
- Daily digest message template
- Archived literature notes
How it works
- Each paper is scored across six weighted dimensions, with per-dimension caps and a recalculated total.
- If Semantic Scholar is unavailable, search falls back to OpenAlex, Crossref, and arXiv.
- The pipeline only writes to the raw literature directory; wiki updates stay manual.
Before installing
The daily cron job runs locally, so the machine must be running at the scheduled time.
More Study tools skills
See all- teach974kTeaches you a new skill or concept step by step inside your current workspace.
- scaffold-exercises554kCreates exercise folders with sections, problems, solutions and explainers that pass linting.
- academy-guide18.9kPoints you to matching Claude Academy courses and tutorials when you ask how to use Claude.
- academy-skill2.2kPulls official Binance Academy lessons, glossary entries and articles to answer crypto questions and build learning plans.
Also in this pack
- nature-figure20.6kCreates, revises and exports submission-ready scientific figures and multi-panel plots in Python or R.
- nature-polishing19.3kPolishes, translates or tightens existing academic prose while keeping facts and terminology intact.
- nature-paper2ppt17.8kBuilds a Chinese academic PPTX from a paper or reading notes, with source figures and speaker notes.
- nature-writing17.4kDrafts or restructures manuscript sections, arguments and initial-submission materials from your evidence.