Scientific Skills for Claude plugin
A set of skills for searching scientific databases, PubMed, arXiv, ChEMBL, ClinicalTrials and more, with plain-language queries.
- 64 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Queries and the Valyu key go to an external paid API.
- On first run Claude asks to paste the key into chat; setting it through an environment variable is safer.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/scientific-skills-for-claudeDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Queries and the Valyu key go to an external paid API.
- On first run Claude asks to paste the key into chat; setting it through an environment variable is safer.
This is third-party code. Review the repository files before installing.
What it does
Scientific Skills for Claude bundles skills that search scientific literature and specialist databases through Valyu semantic search. A query is written in plain words, with no database syntax or parameters. The reply is JSON with a title, link, full article text, a relevance score and figure links. Individual skills cover PubMed, arXiv, bioRxiv, medRxiv, ChEMBL, DrugBank, ClinicalTrials.gov, FDA drug labels, Open Targets and patents. Three aggregated skills combine sources: a general literature review, medical research and drug discovery. On first run Claude asks for a Valyu key and saves it to `~/.valyu/config.json`.
Who it is for. Researchers, clinicians and students who compile literature reviews or look up molecules and clinical trials through Claude Code and other agents.
Good fit when
- A review needs recent papers on a topic from PubMed, arXiv and bioRxiv preprints at once.
- You must find clinical trials and FDA labels for one drug.
- You look for drug targets and their links to diseases.
- The agent needs the full article text with figures, not just the abstract.
- A patent search runs on a natural query without database syntax.
Not a fit when
- A paid third-party service is unacceptable: every query goes through the Valyu API and spends its credits.
- You need exact boolean queries and native database filters: the skills search by meaning.
- Data must not leave the machine, for example with closed research.
Example request
Search PubMed for 2024 papers on CRISPR and recap the three main results with links.Limitations
Node.js 18 or newer and a Valyu key are required: signup gives 10 dollars of free credits, after which search is paid. Queries go to an external API, and the key is kept in `~/.valyu/config.json` or the `VALYU_API_KEY` variable. Filtering works through query wording, the `maxResults` parameter and parsing output with jq. Coverage and freshness are set by the Valyu service, and the MIT license is stated in individual skill files.
How to disable. Remove the plugin with `/plugin uninstall` or delete the skills directory by hand. Delete the Valyu key from `~/.valyu/config.json` and from the environment variable.
FAQ
How do I install Scientific Skills?
Run `npx skills i yorkeccak/scientific-skills`: the skills install into Claude Code, Codex, Cursor, Gemini CLI and other agents. Claude Code also has a plugin: `/plugin install yorkeccak/scientific-skills`.
Where do I get a Valyu key?
At platform.valyu.ai: signup gives 10 dollars of credits. Set the key through the `VALYU_API_KEY` variable or the `~/.valyu/config.json` file. If there is no key, Claude asks for it on the first query.
Which databases are covered?
PubMed, arXiv, bioRxiv, medRxiv, ChEMBL, DrugBank, ClinicalTrials.gov, FDA drug labels, Open Targets and patents. The aggregated skills `literature-search`, `biomedical-search` and `drug-discovery-search` combine several sources.
What does a search return?
JSON with `title`, `url`, `content` holding the full text, `source`, `relevance_score` and an `images` list with the paper's figures. The `maxResults` parameter limits the number of results.
Can I customize search parameters?
The skills are built on plain-language queries. Refine the wording, change `maxResults` and filter the output JSON with jq. There are no native database filters here.
Discussion
No questions or reviews yet.
Related
PluginsA Claude Code skill that turns the agent into a researcher: multi-step source collection, citation verification, and a wiki report that grows session by session
Paper search across five databases with deduplication, citation checks and RIS or BibTeX export for Claude Code and Codex
A Claude Code skill that turns "research this topic" into a multi-phase investigation: hypotheses, parallel search, a file per source and citation checks
Deep Research
Deep Research Skill for Claude Code
A deep research skill for Claude Code: multi-phase search, claim verification and cited reports in Markdown, HTML and PDF
CNKI Skills
CNKI Skills for Claude Code
A Claude Code skill set that drives CNKI through Chrome DevTools MCP: search papers, browse journals, download PDFs and export to Zotero
A skill for Claude Code and Codex: one query goes to 52 search channels and the results merge into one deduplicated JSON
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| Scientific Skills for Claude (this card) | no | 64 | |
| Scientific Agent Skills | no | 56,977 | gives agents scientific workflows and tools; Scientific Skills searches databases with full text through Valyu |
| arxiv-to-prompt | no | 172 | turns an arXiv paper into prompt text; here the search runs across many databases at once |
README in short
The authors stress zero API parameters, semantic search and full text with images. A comparison table sets these skills against keyword search and separate integrations.