@e-labus/mcp

Health Intelligence for AI Assistants - HIPAA-compliant lab analysis, clinical insights, and biomarker interpretation

What It Does

E-Labus turns raw lab results into clinical intelligence. This MCP server gives any AI assistant - Claude, GitHub Copilot, OpenAI Codex, Cursor, Windsurf - the ability to analyze lab reports, interpret biomarkers, and generate physician-grade health narratives.

Hand your AI assistant a list of lab values and it will:

• Flag abnormal biomarkers with clinical context and severity
• Identify cross-biomarker patterns (e.g., metabolic syndrome clusters)
• Score overall health on a 0-100 scale with category breakdowns
• Generate plain-English summaries a patient can understand
• Predict risk across 10 conditions including diabetes, CKD, and cardiovascular disease
• Extract structured data directly from lab PDF files via OCR

No PHI is ever stored in the MCP server itself - all processing happens inside the HIPAA-compliant E-Labus API.

Works With

| AI Tool | Status | |---|---| | Claude Code | Supported | | GitHub Copilot | Supported | | OpenAI Codex CLI | Supported | | Cursor | Supported | | Windsurf | Supported | | Gemini CLI | Supported |

Any MCP-compatible client using the Model Context Protocol specification works with this server.

Quick Start

1. Get an API Key

Free sandbox key - 500 credits, no credit card required: https://e-labus.com/developers

2. Install

npm install @e-labus/mcp

Or use directly with npx - no installation needed.

3. Configure for Your Tool

Claude Code:

claude mcp add e-labus \
  --command "npx @e-labus/mcp" \
  --env ELABUS_API_KEY=your_key_here

Cursor (.cursor/mcp.json):

{
  "mcpServers": {
    "e-labus": {
      "command": "npx",
      "args": ["@e-labus/mcp"],
      "env": { "ELABUS_API_KEY": "your_key_here" }
    }
  }
}

GitHub Copilot (.github/copilot/mcp.json):

{
  "mcpServers": {
    "e-labus": {
      "command": "npx",
      "args": ["@e-labus/mcp"],
      "env": { "ELABUS_API_KEY": "your_key_here" }
    }
  }
}

OpenAI Codex CLI:

export ELABUS_API_KEY=your_key_here
codex mcp add npx @e-labus/mcp

Windsurf (~/.codeium/windsurf/mcp_config.json):

{
  "mcpServers": {
    "e-labus": {
      "command": "npx",
      "args": ["@e-labus/mcp"],
      "env": { "ELABUS_API_KEY": "your_key_here" }
    }
  }
}

Available Tools

| Tool | Description | Credits | Example Use | |---|---|---|---| | analyze_lab_results | Score and interpret biomarker values with clinical context | 5 | "Is this HbA1c concerning?" | | extract_lab_from_pdf | OCR a lab PDF and return structured biomarker data | 10/page | "Extract values from this Quest report" | | normalize_biomarkers | Standardize test names and units across labs | 1/test | "Normalize these lab names" | | generate_clinical_insights | AI pattern analysis with risk tiers and follow-up recommendations | 25 | "What patterns do you see?" | | get_health_score | Aggregate wellness score 0-100 with category breakdown | 5 | "What's this patient's overall score?" | | generate_health_narrative | Plain-English health summary suitable for patients | 20 | "Explain these results to my patient" | | predict_disease_risk | Risk prediction for 10 conditions including diabetes, CKD, and cardiovascular disease | 15 | "What's the diabetes risk?" |

Example Conversations

Example 1 - Routine screening review

User:      Analyze these labs for a 52-year-old male:
           Glucose 118 mg/dL (ref 70-99), HbA1c 6.3% (ref <5.7),
           LDL 152 mg/dL (ref <100), Triglycerides 198 mg/dL (ref <150)

Assistant: [calls analyze_lab_results]

           CLINICAL ANALYSIS - 52-Year-Old Male
           =====================================
           Overall Health Score: 61/100 (Moderate Risk)

           FLAGGED BIOMARKERS (3 of 4 abnormal):

           Glucose: 118 mg/dL  HIGH (ref 70-99)
           - Impaired fasting glucose. Pre-diabetic range.
           - Risk: Moderate. Recommend HbA1c confirmation + dietary review.

           HbA1c: 6.3%  HIGH (ref <5.7%)
           - Confirms pre-diabetes (5.7-6.4% range).
           - Risk: Moderate. Lifestyle intervention recommended now.

           LDL: 152 mg/dL  HIGH (ref <100)
           - Elevated LDL increases cardiovascular risk, compounded
             by pre-diabetic status.
           - Risk: High. Statin therapy discussion warranted.

           Triglycerides: 198 mg/dL  BORDERLINE HIGH (ref <150)
           - Combined with elevated glucose: classic metabolic syndrome cluster.

           PATTERN DETECTED: Metabolic Syndrome (3/5 criteria met)
           RECOMMENDED NEXT STEPS: Fasting insulin, HOMA-IR, liver panel

Example 2 - Patient-facing explanation

User:      Generate a plain-English summary of these results I can
           share directly with my patient.

Assistant: [calls generate_health_narrative]

           Your Lab Results - What They Mean for You
           ==========================================
           Your recent blood work shows some numbers that are slightly
           above the normal range and are worth paying attention to.

           Your blood sugar (glucose) and a longer-term sugar marker
           called HbA1c are both in the pre-diabetes range. This does
           not mean you have diabetes - it means your body is having
           a harder time managing sugar than it should. The good news
           is that lifestyle changes now can often reverse this trend.

           Your cholesterol (LDL) is also elevated. Combined with
           the blood sugar findings, your doctor may want to discuss
           medication options at your next visit.

           Suggested actions: 30 minutes of walking 5 days per week,
           reduce refined carbohydrates and sugars, follow up in 3 months.

Example 3 - PDF extraction

User:      Here is a Quest Diagnostics PDF. Extract all the lab values.

Assistant: [calls extract_lab_from_pdf with the uploaded file]

           Extracted 14 biomarkers from 2 pages:

           CBC:
           - WBC: 7.2 K/uL (ref 4.5-11.0) - Normal
           - RBC: 4.8 M/uL (ref 4.5-5.9)  - Normal
           - Hemoglobin: 14.1 g/dL (ref 13.5-17.5) - Normal
           - Hematocrit: 42% (ref 41-53) - Normal
           - Platelets: 248 K/uL (ref 150-400) - Normal

           Metabolic Panel:
           - Glucose: 118 mg/dL (ref 70-99) - HIGH
           - BUN: 18 mg/dL (ref 7-25) - Normal
           - Creatinine: 0.9 mg/dL (ref 0.6-1.2) - Normal
           - eGFR: >60 - Normal

           Lipid Panel:
           - Total Cholesterol: 228 mg/dL (ref <200) - HIGH
           - LDL: 152 mg/dL (ref <100) - HIGH
           - HDL: 42 mg/dL (ref >40) - Normal
           - Triglycerides: 198 mg/dL (ref <150) - BORDERLINE

           Thyroid:
           - TSH: 2.1 mIU/L (ref 0.4-4.0) - Normal

Pricing

| Plan | Credits/Month | Price | HIPAA BAA | |---|---|---|---| | Sandbox | 500 | Free | Not included | | Starter | 5,000 | $49/mo | Included | | Growth | 25,000 | $199/mo | Included | | Scale | 150,000 | $799/mo | Included | | Enterprise | Unlimited | Custom | Included + MSA |

Credits never expire. Unused credits roll over within your plan period.

Get started at e-labus.com/developers

HIPAA & Security

• BAA included on all paid plans - Business Associate Agreement available from the developer dashboard on Starter and above
• HTTPS only - all traffic encrypted in transit with TLS 1.2+
• AES-256 encryption at rest - all data encrypted at the storage layer
• API keys are SHA-256 hashed before storage - raw keys are never persisted
• Zero PHI stored in the MCP server itself - the MCP server is a thin proxy; all data stays in E-Labus infrastructure
• Data residency: US - AWS us-east-1
• SOC 2 Type II - in progress (available on Enterprise)

Requirements

• Node.js 18 or higher
• An E-Labus API key (get one free)

License

MIT - see LICENSE

Docs - Dashboard - GitHub - Support