Biolit

Install

Works with Claude Code, Codex, VS Code, Cursor, Windsurf, Claude Desktop. Read the repository before you give any tool access to your data.

npx agentmarket install connector/rachadele-biolit

About

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this server can do and what it asks for, from its tool list and install config. Checked Oct 25, 2026.

Tools

10 read · 2 write · 0 delete

Secrets

3 (API keys, tokens or passwords)

Install scripts

Not scanned yet

Runs as

Python package

Tools (12)

Read from the source code; the running server may differ. Descriptions are the server’s own words; hidden characters are shown as [U+…].

• extract_fields

  Extract structured fields from a paper using the LLM. Args: title: Paper title. text: Abstract or full text to extract from. fields: Comma-separated field names, e.g. "methodology, sample_type, causal_claims, summary" mesh_terms: Comma-separated MeSH terms (optional). Returns: JSON object with one key per requested field, plus title, url, pmid, doi, and text_source keys.

• fetch_fulltext

  Retrieve full text for a PubMed ID. Tries sources in order: PMC JATS XML → Europe PMC JATS XML → preprint XML (bioRxiv/medRxiv) → Unpaywall PDF → Semantic Scholar open-access PDF → abstract fallback. Args: pmid: PubMed ID. unpaywall_email: Your email for the Unpaywall API. Falls back to the UNPAYWALL_EMAIL env var. Required only for the Unpaywall step. sections: Comma-separated section names to include, e.g. "methods,results". Default: all sections. Returns: {"text": "...", "source": "pmc_fulltext" | "europepmc_fulltext" | "prep…

• fetch_geo_fulltext

  Retrieve full text for a GEO accession via its linked PubMed IDs. Fetches the GEO record, then tries each linked PMID using the same full-text chain as fetch_fulltext (PMC → Europe PMC → Unpaywall → S2). Falls back to the linked paper's abstract, then to the GEO metadata text. Args: accession: GEO accession, e.g. "GSE53987". unpaywall_email: Your email for the Unpaywall API. Falls back to the UNPAYWALL_EMAIL env var. sections: Comma-separated section names to include, e.g. "methods,results". Default: all sections. Returns: {"text": "...", "source": "geo_linke…

• fetch_geo_record

  Fetch and parse a GEO series record. Accepts GSE, GDS, GSM, or GPL accessions. Returns title, summary, overall design, experiment type, organism(s), and any linked PMIDs. The 'abstract' key contains a combined text blob suitable for passing to screen_paper or extract_fields.

• fetch_pubmed_metadata

  Fetch metadata for a PubMed ID. Returns title, abstract, MeSH terms, DOI, and a PubMed URL. Use this before calling screen_paper or extract_fields so you have the paper text and metadata to pass along.

• fetch_supplementary

  Fetch a paper's supplementary files (supplementary methods, tables) from Europe PMC. Supplementary methods often carry curation-relevant detail (strain backgrounds, cell-line provenance, reagents) that the main article body omits. The full-text fetchers never return it because the JATS only references supplementary files; this retrieves and parses them. Provide one of pmcid / pmid / doi (pmcid preferred). Open access only — paywalled supplements are not retrievable. Args: pmcid: PubMed Central ID (e.g. "PMC10700610"), preferred. pmid: PubMed ID; resolved to a PMCID if pmcid is absen…

• get_version

  Return the installed biolit package version.

• lookup_s2_pdf

  Look up whether Semantic Scholar has an open-access PDF for a DOI. Useful for checking PDF availability before committing to a full fetch, or for retrieving the PDF URL to pass to another tool. Args: doi: DOI string, e.g. "10.1101/2021.11.01.466731". Returns: {"doi": "...", "pdf_url": "..." | null, "available": true | false}

• read_pmids_from_eml

  Parse PMIDs out of a PubMed alert .eml file. Args: eml_path: Absolute path to the .eml file on disk. Returns: {"pmids": ["12345678", ...], "count": N}

• resolve_doi

  Resolve a DOI to a PMID and/or PMCID via the NCBI ID Converter. Useful for chaining with other tools when you have a DOI but need a PMID. Args: doi: A DOI string, e.g. "10.1038/s41588-021-00974-7". Returns: {"doi": "...", "pmid": "..." | null, "pmcid": "..." | null}

• run_pipeline

  Run the fetch → (screen) → (extract) pipeline on a mixed list of identifiers and write a CSV. Accepts PMIDs, DOIs, and GEO accessions in any combination. Each identifier is auto-detected and routed to the appropriate fetcher. Full-text retrieval is attempted for all record types — GEO records use linked PMIDs as the source, falling back to GEO metadata if none are accessible. This is equivalent to running `biolit --ids ...` from the command line. Args: ids: Comma-separated identifiers — PMIDs, DOIs, GEO accessions, or any mix. Example: "41795042,GSE53987,10.1101/2025.03.17.25324…

• screen_paper

  Ask the LLM whether a paper meets a relevance criterion. Args: title: Paper title. text: Abstract or full text to screen. criterion: A yes/no question about relevance, e.g. "Is this paper specifically about schizophrenia genomics?" mesh_terms: Comma-separated MeSH terms (optional but improves accuracy). Returns: {"relevant": true | false, "reason": "one sentence"}

Signals

Stars

2

Forks

0

Open issues

1

Weekly downloads

118

Last commit

Sep 23, 2026 (1 month ago)

Last release

Apr 1, 2026 (7 months ago)

Last package publish

Aug 19, 2026 (2 months ago)

From GitHub (fetched Oct 24, 2026) and the package registry.

Details

Registry name

io.github.rachadele/biolit

Version

0.1.25

Auth

API key

Packages

pypi biolit

License

MIT

Repository

github.com/rachadele/biolit

First seen

Oct 24, 2026

Last checked

Oct 24, 2026

Source

Official MCP Registrygithub-graphql