Skip to main content
Glama

Doctor Claude

An MCP server that:

  1. Provides Claude Code with access to peer-reviewed medical information from trusted sources: MedlinePlus and StatPearls (NCBI).

  2. Acts as an unliscensed diagnostic medical expert.

Installation

  1. Clone and navigate to the repository:

git clone https://github.com/gholtzap/doctor-claude.git
cd doctor-claude
  1. Run the setup script:

./setup.sh

Manual Setup

  1. Clone and navigate to the repository:

git clone https://github.com/gholtzap/doctor-claude.git
cd doctor-claude
  1. Install dependencies and build:

npm install
npm run build
  1. Add to Claude Code (must be run from the repository directory):

claude mcp add --transport stdio doctor-claude -- node $(pwd)/build/index.js

Important: Make sure you're in the doctor-claude directory when running the above command.

Next Steps

  1. (Optional) Set your patient profile (see below)

  2. Launch Claude Code and type /doctor-claude:diagnostic_consultation

Related MCP server: StatPearls MCP Server

Patient Profile

Create a patient-profile.json file in your working directory to provide personalized medical context during consultations. This file is automatically loaded when you start a diagnostic consultation.

Example file (patient-profile.json):

{
  "age": 45,
  "sex": "female",
  "weight": { "value": 150, "unit": "lbs" },
  "height": { "value": 65, "unit": "in" },
  "chronicConditions": ["Type 2 Diabetes", "Hypertension"],
  "medications": ["Metformin 500mg twice daily", "Lisinopril 10mg daily"],
  "allergies": ["Penicillin", "Shellfish"],
  "surgicalHistory": ["Appendectomy (2015)"],
  "familyHistory": ["Mother: Breast cancer", "Father: Heart disease"]
}

All fields are optional. Copy patient-profile.example.json to patient-profile.json and customize it.

Privacy: Your profile has the same privacy level as sending it in a claude chat. I, the developer, have 0 access to any profiles, as they are stored locally on your machine.

Prompts

diagnostic_consultation

Engages Claude in a systematic medical consultation process. Automatically includes your patient profile if available.

Important: This is for educational purposes only. If you are feeling unwell, see a doctor.

Usage in Claude Code:

/doctor-claude:diagnostic_consultation

Available Tools

6 tools
calculate_clinical_scoreA

Calculate clinical decision rule scores to help assess disease severity, risk stratification, and guide clinical decision-making. Supports multiple evidence-based calculators: CURB-65 (pneumonia severity), Centor Score (strep throat probability), Wells DVT (deep vein thrombosis risk), Wells PE (pulmonary embolism risk), HEART Score (chest pain cardiac risk), CHA2DS2-VASc (stroke risk in atrial fibrillation), GCS (Glasgow Coma Scale for consciousness assessment), qSOFA (sepsis screening), Alvarado Score (appendicitis), Glasgow-Blatchford Score (upper GI bleeding risk), NIHSS (NIH Stroke Scale for stroke severity), SOFA (Sequential Organ Failure Assessment), PERC (Pulmonary Embolism Rule-out Criteria), TIMI (Thrombolysis in Myocardial Infarction Risk Score), MELD (Model for End-Stage Liver Disease), GAD-7 (Generalized Anxiety Disorder-7). These tools help determine appropriate level of care (outpatient vs hospital) and testing strategies. This tool provides EDUCATIONAL information only - all clinical decisions must be made by qualified healthcare providers.

Inputs: inputs (required), calculator (required).

delete_patient_profileB

Delete the saved patient profile information.

Inputs: no named parameters.

fetch_medical_articleA

Fetch and parse the full content of a medical article from MedlinePlus or StatPearls. Only URLs from medlineplus.gov and ncbi.nlm.nih.gov are allowed for security. This tool provides EDUCATIONAL information only - not medical advice, diagnoses, or treatment recommendations.

Inputs: url (required).

get_patient_profileA

Retrieve the saved patient profile information. Returns null if no profile has been saved.

Inputs: no named parameters.

search_medical_infoA

Search for peer-reviewed medical information from MedlinePlus and StatPearls (NCBI). Returns a list of relevant articles with titles, URLs, and descriptions. IMPORTANT: Use simple, focused queries with 1-3 key medical terms for best results. Avoid overly complex multi-word queries. If searching for specific patterns or presentations, use the core condition/topic (e.g., "child abuse burns" instead of "geometric patterned skin lesions burns infants child abuse"). The tool will automatically try simpler queries if the initial search returns no results. This tool provides EDUCATIONAL information only - not medical advice, diagnoses, or treatment recommendations.

Inputs: query (required), source.

set_patient_profileA

Save patient profile information (age, weight, height, chronic conditions, medications, allergies, etc.). This information is stored locally and will be used to provide more personalized medical information during consultations. All fields are optional - you can provide as much or as little information as you want.

Inputs: age, sex, height, weight, +5 more.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A4.2/5.0

    Scored across 6 tools

    Disambiguation5/5

    Each tool targets a distinct operation: searching for articles, fetching article content, managing patient profile (set/get/delete), and calculating clinical scores. No two tools overlap in purpose.

    Naming Consistency5/5

    All tool names follow a consistent `verb_noun` pattern with snake_case (e.g., `fetch_medical_article`, `get_patient_profile`), making the action and resource clear.

    Tool Count5/5

    With 6 tools, the server is well-scoped: 3 for patient profile management, 2 for medical information retrieval, and 1 for clinical scoring. This covers core functionality without unnecessary bloat.

    Completeness5/5

    The tool set covers the full lifecycle: search/fetch medical info, save/retrieve/delete patient context, and calculate relevant clinical scores. No obvious missing operations for the stated purpose.

    Maintenance

    ActivityInactive
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      D
      maintenance
      A specialized Model Context Protocol server that enhances AI-assisted medical learning by connecting Claude Desktop to PubMed, NCBI Bookshelf, and user documents for searching, retrieving, and analyzing medical education content.
      7
      MIT
    • F
      license
      Not graded
      quality
      D
      maintenance
      Provides AI systems with access to peer-reviewed medical information from StatPearls, enabling searches for diseases and medical conditions with comprehensive, reliable content formatted in AI-friendly Markdown.
      2
      -
    • A
      license
      Not graded
      quality
      F
      maintenance
      Provides comprehensive medical information by querying authoritative sources including FDA drug database, WHO health statistics, PubMed literature, RxNorm nomenclature, Google Scholar, and Australia's PBS API through 22+ specialized tools.
      83 npm
      3
      MIT
    • A
      license
      A
      quality
      C
      maintenance
      Enables searching medical literature, clinical trials, drug information, guidelines, and medical fee schedules directly from Claude, returning only original sources and text without medical judgment.
      9
      MIT