Planned support for custom SQLite data sources, allowing drug and molecule data to be queried from local SQLite databases alongside CSV and Excel sources.
Planned integration for Wikidata identifier cross-references, enabling drug and molecule identifier lookups across sources.
Planned integration for retrieving Wikipedia pharmacokinetics information for drugs and molecules.
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
MCP servers and a Python library for looking up drug and molecule data — identifiers, measured and computed properties, and drug information — across public databases, with the source recorded on every value.
Status: early development. Nothing is released yet. See the roadmap below.
Small molecules, peptides (including modified and cyclic peptides via HELM), and conjugates such as antibody–drug conjugates (ADCs). Oligonucleotides are planned.
Identifiers: InChIKey, SMILES, InChI, CAS, PubChem CID, ChEMBL ID, UNII, Wikidata QID, sequence, HELM.
Related MCP server: PubChem MCP Server
Package | Status | What it does |
| In progress | Shared identifiers, data models, units, HTTP client |
| Planned | Identifier resolution; computed and experimental properties |
| Planned | Drugs, mechanisms, targets, bioactivity, ADC components |
| Planned | Identifier cross-references; Wikipedia pharmacokinetics |
Also planned: configurable unit and source-preference profiles, custom data sources (CSV, Excel, SQLite), a REST API, and a combined server.
Architecture decisions are recorded in docs/decisions, and requirements in docs/requirements.
pharmakon's code is Apache-2.0. Data it retrieves remains under each source's terms:
PubChem (NCBI) — public domain data; see the PubChem usage policy.
ChEMBL (EMBL-EBI) — CC BY-SA 3.0; attribution required.
Wikidata — CC0. Wikipedia text — CC BY-SA 4.0.
Please respect each service's rate limits; pharmakon enforces them by default.
For research use only. Not for clinical or diagnostic decisions. Values come from third-party sources and may be incomplete or wrong; always check the cited source.
Apache-2.0. Copyright 2026 Andrew Stuart.
This server cannot be deployed
Link compounds to protein targets, rank bioactivity, and look up drug mechanisms and indications.
Search PubChem compounds, properties, safety data, bioactivity, and cross-references.
MyChem.info MCP. Aggregated drug / chemical-compound annotation from the BioThings API: ChEMBL…
Linked life-science search: 43M+ papers, 1.2M+ trials, drugs, genes, FDA/EMA approvals, patents.