Pharmacy MCP Server
Enables medication reference workflows including drug search, dosing calculations, Taiwan regulatory lookups, hospital prescription helpers, and educational PK/DDI simulation.
README
Pharmacy MCP Server
Pharmacy MCP is a Python MCP server for medication reference workflows. It combines drug search, label summaries, dosing calculators, Taiwan TFDA/NHI lookups, hospital prescription helpers, and a trusted PBPK-lite formula catalog for educational PK/DDI simulation.
Traditional Chinese: README.zh-TW.md
What Is New In 0.9.1
- Hardened interaction outputs so local database guidance is surfaced as non-prescriptive management considerations with safety metadata.
- Added fail-closed simulation guardrails for NaN, infinity, unstable denominators, and catalog metadata drift.
- Added release artifact auditing and packaged-wheel smoke tests.
- Made Streamable HTTP ASGI exports lazy and mount-aware.
What Is New In 0.9.0
- Built-in trusted PK/DDI formula catalog under
src/pharmacy_mcp/data/formulas/. - Deterministic simulation service for one-compartment concentration, repeated-dose accumulation, renal clearance adjustment, CYP reversible inhibition, AUC ratio, and temperature-corrected elimination.
- Mechanism-aware DDI tools for supported CYP inhibition examples such as warfarin/fluconazole and statin/clarithromycin.
- Modern MCP surfaces: tools, read-only resources, resource templates, prompts, Streamable HTTP deployment helpers, and structured outputs.
- Release gates for pytest coverage, ruff, mypy, bandit, package build, and artifact audit.
Simulation outputs are screening estimates only. They always include the project
disclaimer and not_for_direct_clinical_decision: true.
Install
git clone https://github.com/u9401066/pharmacy-mcp.git
cd pharmacy-mcp
uv sync --all-extras
Run
# stdio transport for local MCP clients
uv run pharmacy-mcp
# explicit stdio
uv run pharmacy-mcp --transport stdio
# Streamable HTTP
uv run pharmacy-mcp --transport streamable-http --host 0.0.0.0 --port 8000
# ASGI app
uv run uvicorn pharmacy_mcp.presentation.server:app --host 0.0.0.0 --port 8000
Environment variables use the PHARMACY_MCP_ prefix. Supported settings include
PHARMACY_MCP_TRANSPORT, PHARMACY_MCP_HOST, PHARMACY_MCP_PORT,
PHARMACY_MCP_MOUNT_PATH, PHARMACY_MCP_STREAMABLE_HTTP_PATH, and
PHARMACY_MCP_STATELESS_HTTP.
Claude Desktop Example
{
"mcpServers": {
"pharmacy": {
"command": "uv",
"args": ["run", "pharmacy-mcp", "--transport", "stdio"],
"cwd": "/path/to/pharmacy-mcp"
}
}
}
MCP Tools
Drug reference:
search_drugget_drug_infoget_drug_dosageget_drug_warningscheck_drug_interactioncheck_multi_drug_interactionscheck_food_drug_interaction
Dosing calculators:
calculate_dose_by_weightcalculate_dose_by_bsacalculate_creatinine_clearancecalculate_pediatric_dosecalculate_infusion_rateconvert_dose_units
Taiwan TFDA/NHI:
search_tfda_drugget_nhi_coverageget_nhi_drug_pricetranslate_drug_namelist_prior_authorization_drugslist_nhi_coverage_rules
Hospital prescription workflow:
get_formulary_itemsearch_formularyget_renal_adjustmentvalidate_ordersubmit_orderstop_order
Trusted formula and simulation:
list_formula_catalogget_formula_detailsexplain_interaction_mechanismsimulate_pk_interactionsimulate_concentration_time
MCP Resources And Prompts
Resources:
pharmacy://server/disclaimerpharmacy://formulaspharmacy://formulas/{formula_id}pharmacy://validation/formulas
Prompts:
ddi_analysis_workflowformula_review_checklist
Trusted Formula Model
Formula metadata is committed as data, not executed as arbitrary runtime code. Each trusted formula has an ID, expression, parameters, units, assumptions, limitations, references, and validation cases. The Python simulation service dispatches by formula ID to reviewed implementations.
External formula generators such as NSForge can be used as companion authoring tools, but Pharmacy MCP does not vendor NSForge in 0.9.x. Generated formulas must be reviewed, committed to the trusted catalog, and covered by tests before production tools can use them.
Data Sources
| Source | Provider | Data |
|---|---|---|
| RxNorm API | NIH/NLM | Drug names and concepts |
| openFDA | FDA | Drug labels and warnings |
| DailyMed | NLM | Label sections |
| TFDA Open Data | Taiwan TFDA | Taiwan permits and drug names |
| NHI Open Data | Taiwan NHI | Coverage rules and reimbursement data |
| Local catalog | Pharmacy MCP | Trusted PK/DDI formulas and hospital mock data |
Architecture
src/pharmacy_mcp/
domain/ Entities and value objects
application/services/ Use-case services and simulation orchestration
infrastructure/api/ External API clients
infrastructure/cache/ Disk cache adapter
infrastructure/knowledge/Local formularies and trusted formula catalog
presentation/ FastMCP server, tools, resources, prompts
Verification
uv run pytest --cov=src --cov-report=term-missing
uv run ruff check src tests
uv run mypy src
uv run bandit -q -r src
uv build
uv run python scripts/audit_release_artifacts.py dist
Disclaimer
This project is for reference, education, and workflow support only. It does not provide medical advice and must not be used as the sole basis for clinical decisions. Consult qualified healthcare professionals and validated clinical systems for patient care.
License
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.