Preprint Fulltext MCP server

preprint-fulltext

English | 简体中文 | 繁體中文 | 한국어 | Deutsch | Español | Français | Italiano | 日本語

Retrieve the full text of bioRxiv / medRxiv / arXiv preprints as clean, structured, embedding-ready data — from a CLI, a Python library, or an MCP server.

preprint-fulltext turns a DOI (or a search) into structured sections (abstract / introduction / methods / results / discussion), a single JSON/Markdown document, or a chunked JSONL/Parquet corpus ready for embeddings and RAG. openRxiv text-and-data-mining (TDM) compliance is enforced structurally, not left to the user.

**"Embedding-ready" means the output is clean, section-aware, token-bounded chunks —

ready to feed to your embedding model. Computing embeddings is an optional last step

you own; this tool does not bundle an embedding model.**

Contents

• Why · Who it's for · Full text for AI-driven science
• Features · Install · Quickstart (CLI) · Typical workflows
• Python library · MCP server (coding agents) · Data sources & routing
• Configuration · Compliance · Development · Contact · License

Why

Preprint full text is scattered across incompatible channels: Europe PMC serves JATS XML for the open-access subset, the openRxiv S3 buckets hold the authoritative .meca corpus (requester-pays), OpenAlex is a catalog with n-gram-only full-text search, and the bioRxiv/medRxiv websites render HTML. preprint-fulltext unifies them behind one canonical data model and one shared JATS parser, so you get the same structured output no matter where a document came from.

Who it's for

preprint literature.

• ML / NLP researchers building embedding corpora or RAG systems over the

analysis, extraction, or LLM pipelines.

• Bioinformaticians and labs who need a paper's methods/results as clean text for

full text or search the literature mid-task.

• Coding agents (via the MCP server / SKILL.md) that need to pull a preprint's

scraping HTML.

• Anyone who wants one preprint's sections from a DOI without hand-parsing JATS or

Full text for AI-driven science

Language models and agents reason far more reliably over a paper's methods and results than over its abstract alone — most scientific claims, protocols, quantities, and caveats live in the body. preprint-fulltext gives Claude, Codex, and other agents that body as clean, section-labeled, provenance- and license-tagged text, which is the substrate for grounded scientific reasoning and deep research:

• Literature deep-research — read across many papers' full text, not just abstracts.
• Methods / protocol extraction — pull exact procedures, parameters, and datasets.
• Claim verification — check a stated result against the actual Results section.
• Reproducibility & meta-analysis — compare methods and numbers across studies.
• RAG over your own corpus — section-aware, token-bounded chunks with citations.

Shortened. The full README is on GitHub.