Skip to main content
Glama

MedOrqis — Medical AI Skills Ecosystem (v2.1.2)

Live Interactive Catalog DOI PyPI version Glama MCP Server TDQS Score M8ven Score Hugging Face Space HF Model HF Dataset Dedicated Model Repo HF MCP Server Skills CLI npm version Verification: 36/36 Passed (100%) GitHub CI License: Apache-2.0

MedOrqis is an open-source, evidence-grounded, cross-platform ecosystem of 24 specialized clinical AI skills, 12 deterministic zero-token clinical calculators & safety engines, and multi-platform adapters (Skills CLI, Claude Code, Gemini CLI, Model Context Protocol, and OpenAI).

TIP

🌐Interactive Web Hub & Live Zero-Token Calculators:
Explore the live searchable 24-skill catalog, test interactive clinical calculators (ESC 2024 CHA₂DS₂-VA, CKD-EPI 2021, Bedside Schwartz, HAS-BLED, NEWS2, DDI Checker), and inspect dialect benchmarks directly in your browser at https://seotarek.github.io/MedOrqis/.

⚡ Fleet Install in 1-Second (Claude Code, Cursor, Codex, Antigravity, OpenCode):

npx skills add seotarek/MedOrqis

Built with an evidence-first, patient-safety-first architecture, MedOrqis eliminates medical hallucination risks in high-stakes clinical calculations by executing arithmetic, unit conversions, PHI scrubbing, negation parsing, and drug interaction screenings through a local, deterministic Python core with zero LLM API token overhead.


⚡ 1-Minute Quick Start & Integration Options

🛠️ Option 1: One-Command Fleet Install (Skills CLI)

Install all 24 skills directly into your active coding agent workspace (Claude Code, Cursor, Antigravity, OpenCode, Codex):

# Install the complete MedOrqis ecosystem (all 24 skills)
npx skills add seotarek/MedOrqis

# Or install specific clinical skills:
npx skills add seotarek/MedOrqis --skill emergency-triage-redflags
npx skills add seotarek/MedOrqis --skill pharmacology-dosing-interactions
npx skills add seotarek/MedOrqis --skill pediatrics-neonatology

⭐ Support Open-Source Clinical AI: If MedOrqis helps your workflow or coding agents, please give the repo a Star on GitHub — it helps the project reach more clinicians and medical AI engineers!


📦 Option 2: NPX & NPM Interactive Terminal Package (medorqis)

Explore the catalog and copy prompts directly in your terminal:

# Run interactive CLI browser
npx medorqis

# List all 24 skills grouped by safety tier
npx medorqis list

# Print copyable Universal AI Prompt for web chat models
npx medorqis prompt

# Print MCP configuration JSON block
npx medorqis mcp

Or install globally or as a project dependency:

npm install -g medorqis
# or in Node.js apps:
npm install medorqis

⚡ Option 3: Universal AI Model Prompting (Zero Install — Any Model)

You do not need to install anything or run code to use MedOrqis with LLMs! You can send the repository link and prompt directly to ChatGPT (with web access), Claude 3.5 Sonnet, Google Gemini, Cursor, Grok, or Antigravity:

Copy & Paste into any AI model:

Read the open-source clinical AI skills ecosystem MedOrqis at:
https://github.com/seotarek/MedOrqis (Web catalog: https://seotarek.github.io/MedOrqis/)

You are now operating under MedOrqis clinical governance rules (v2.1.1):
1. Enforce zero-hallucination deterministic standards for all medical risk scores and formulas (CHA2DS2-VA ESC 2024, CKD-EPI 2021, Bedside Schwartz, Cockcroft-Gault, Wells DVT/PE, CURB-65, HAS-BLED, NEWS2, Pediatric Dose Capping).
2. Epistemic provenance: Never recommend high-risk interventions based on unconfirmed model hypotheses.
3. Missing Age Gatekeeper: If patient age is unrecorded, do NOT assume a default adult age (30); immediately flag SAFE_ABSTENTION_DATA_DEFICIT.
4. Sentinel Emergency Red Flags: Immediate emergency redirection on acute stroke (FAST), ACS crushing chest pain, anaphylaxis, septic shock, or anticoagulation + active bleeding.
5. Provide evidence-grounded differential diagnosis, drug interaction screening, and red-flag triage across all 24 MedOrqis clinical domains.

Please summarize your activated clinical skills and await clinical input.

🔌 Option 4: Model Context Protocol (MCP Server)

TDQS Score

🛡️ Verified TDQS Grade A · 4.9 / 5.0 (Coherence 5.0 / 5.0): Evaluated for tool disambiguation, parameter semantic boundaries, and schema completeness. View Full TDQS Report.

A. Remote Hosted MCP (Hugging Face ZeroGPU — Zero Installation)

Connect your MCP client directly to our live hosted server on Hugging Face Spaces:

{
  "mcpServers": {
    "medorqis": {
      "url": "https://seotarek-medorqis-mcp-server.hf.space/gradio_api/mcp/sse"
    }
  }
}

B. Local Stdio MCP

Run the local MCP stdio server via the PyPI package:

medorqis-mcp

Or configure Claude Desktop (claude_desktop_config.json):

{
  "mcpServers": {
    "medorqis": {
      "command": "python3",
      "args": ["/path/to/MedOrqis/mcp_server/medorqis_mcp_server.py"]
    }
  }
}

🐍 Option 5: Python Package & Developer Setup

Install directly from PyPI:

pip install medorqis

Or clone for local development:

git clone https://github.com/seotarek/MedOrqis.git
cd MedOrqis
python3 -m pip install -e .

Related MCP server: Medical MCP Server

🧠 Clinical Brain: MedOrqis-LLM

HF Model HF Dataset Dedicated Model Repo

MedOrqis-LLM is the dedicated bilingual (Arabic & English) clinical decision intelligence foundation model developed for the MedOrqis ecosystem. Fine-tuned on top of Applied-Innovation-Center/Karnak-6B-v1.0 (Qwen3 architecture) via QLoRA 4-bit NF4, it acts as the Clinical Brain of MedOrqis:

  • Clinical Narrative & Dialect Comprehension: Accurately parses complex patient histories, lab reports, and colloquial complaints in Modern Standard Arabic, Medical English, and regional dialects (Egyptian & Gulf colloquialisms mapped to ICD-10 and SNOMED-CT).

  • Native Deterministic Tool-Calling: The model is trained never to guess clinical scores, drug dosages, or drug-drug interactions. Instead, it emits structured <tool_call> requests to MedOrqis's verified zero-token Python calculators.

  • Dedicated Model Repository: For serving, inference pipelines (vLLM / Transformers), and fine-tuning weights, visit the standalone repository: 🔗 seotarek/medorqis-llm.


🧮 Pure Deterministic Clinical Calculators (core/clinical_calculators.py)

All medical calculations in MedOrqis execute with zero LLM API token overhead and strict mathematical determinism via pure Python functions in core/clinical_calculators.py:

#

Clinical Calculator

Function Name

Clinical Guideline & Standard

Target Population

Safety Gate & Boundary Enforcement

1

CHA₂DS₂-VA

calculate_cha2ds2_va_esc2024

ESC Guidelines 2024 (Sex-Neutral)

Adults ($\ge 18\text{y}$) with AF

Refuses if Age missing (SAFE_ABSTENTION_DATA_DEFICIT). Removes 1-point female bias.

2

CHA₂DS₂-VASc

calculate_cha2ds2_vasc_aha2023

ACC/AHA/ACCP/HRS 2023

Adults ($\ge 18\text{y}$) with AF

Historical comparative baseline for US guidelines ($+1$ for female sex).

3

CKD-EPI 2021

calculate_ckd_epi_2021_egfr

NKF-ASN Task Force / NEJM 2021

Adults ($\ge 18\text{y}$)

Race-free adult eGFR formula. Hard Refusal if Age $< 18$ years (SAFE_REFUSAL_PEDIATRIC).

4

Bedside Schwartz

calculate_bedside_schwartz_egfr

Schwartz GJ et al. (JASN 2009)

Pediatrics ($< 18\text{y}$)

Height-based eGFR ($0.413 \times \text{Ht}/\text{SCr}$). Hard Refusal if Age $\ge 18$ years (SAFE_REFUSAL_ADULT).

5

Cockcroft-Gault

calculate_cockcroft_gault_crcl

FDA Renal Dosing Standards

Adults ($\ge 18\text{y}$)

Automatically calculates IBW and Adjusted Body Weight (AdjBW $0.4$) if BMI $\ge 30$.

6

HAS-BLED

calculate_has_bled_score

Pisters R et al. (Chest 2010)

Adults on Anticoagulation

Flags High Bleeding Risk ($\ge 3$); warns against unmonitored DOAC initiation.

7

CURB-65

calculate_curb65_score

British Thoracic Society (BTS)

Adults with CAP

Stratifies 30-day mortality risk (Outpatient vs Inpatient vs ICU admission).

8

NEWS2

calculate_news2_score

Royal College of Physicians (2017)

Acute Inpatient Care

Physiological aggregate triage across 6 vital signs; automated sepsis escalation.

9

Wells DVT Criteria

calculate_wells_dvt_score

Wells PS et al. (NEJM 2003)

Suspected DVT

Pre-test probability (Unlikely vs Likely); dictates D-dimer vs Compression Ultrasound.

10

Wells PE Criteria

calculate_wells_pe_score

Wells PS et al. (Ann Intern Med 2001)

Suspected PE

Pre-test probability (Unlikely vs Likely); dictates D-dimer vs CT Pulmonary Angiography.

(Auxiliary dosing & safety engines: calculate_pediatric_weight_dose & calculate_mosteller_bsa in core/clinical_calculators.py, and Contraindicated DDI Screening in core/safety_guardrails.py).


🏥 The 24-Skill Risk-Stratified Taxonomy

Tier 1: Governance, Research & Operations (Class A)

  • medorqis-orchestrator: Context envelope routing and conflict resolution.

  • ebm-research-biostatistics: PICO, CASP, RoB-2 appraisal, and GRADE certainty.

  • medical-documentation-translation: SOAP, SBAR, and BiDi multilingual medical translation.

  • healthcare-ops-telemedicine-ehr: HL7 FHIR mapping and telemedicine intake.

  • health-cybersecurity-privacy: PHI scrubbing and prompt injection defense.

  • public-health-epidemiology: Screening metrics (sens/spec) and IPC protocols.

Tier 2: Diagnostic & Chronic Care Support (Class B)

  • differential-diagnosis-cds: VINDICATE differential framework.

  • laboratory-medicine-pathology: Critical panic values and LOINC mapping.

  • radiology-medical-imaging: ACR Appropriateness Criteria and radiation safety.

  • internal-medicine-chronic: Multi-morbidity chronic disease care plans.

  • infectious-diseases-stewardship: Antimicrobial stewardship and resistance awareness.

  • nursing-clinical-care: NEWS2 early warning score and Braden scale.

  • clinical-nutrition-dietetics: Caloric requirements and food-drug interactions.

  • rehabilitation-physical-therapy: Post-op and stroke rehab protocols.

Tier 3: High-Risk, Emergency & Vulnerable Populations (Class C)

  • emergency-triage-redflags: ABCDE primary survey and emergency redirection.

  • clinical-safety-auditor: Pre-output verification and safety gate check.

  • pharmacology-dosing-interactions: Renal/hepatic adjustments and DDI checks.

  • clinical-toxicology-overdose: Toxidromes, antidotes, and Poison Control escalation.

  • pediatrics-neonatology: Weight-based dose capping and pediatric safety.

  • obstetrics-gynecology: Maternal-fetal safety and pre-eclampsia triage.

  • geriatrics-polypharmacy: 2023 Beers Criteria and STOPP/START.

  • mental-health-psychiatry: Suicide crisis escalation and psychotropic monitoring.

  • perioperative-surgical-care: RCRI pre-op risk and anticoagulation bridging.

  • oncology-palliative-care: Oncologic emergencies and palliative pain control.


🌐 Multilingual & Dialect Regression Benchmark

MedOrqis includes a deterministic natural language processing normalizer verified across 126 curated synthetic test fixtures:

  • 100 Syntactic Negation Sentences: Handcrafted fixtures verifying regex boundary detection and clause-stop logic across 8 dialects and languages: Modern Standard Arabic (ar-MSA), Egyptian (ar-EG), Gulf (ar-GULF), Levantine (ar-LEV), Maghrebi (ar-MAG), English (en-US/GB), French (fr-FR), and Spanish (es-ES).

  • 10 BiDi & Unit Token Strings: Numerical and unit strings verifying bidirectional layout preservation (e.g. Arabic RTL mixing English numbers/units like "500 mg").

  • 16 Trade Brand-to-INN Mappings: Lookup fixtures mapping regional commercial trade brand names to generic International Nonproprietary Names (e.g. Panadol $\to$ Paracetamol).

CAUTION

Two-Axis Governance Separation:

  • Automated Syntactic Unit-Test Pass Rate: 100.0% (126/126 passed).

  • Bedside Real-World Clinical Validation: 0.0% (UNVALIDATED IN-VIVO / PENDING CLINICAL TRIALS).

See full breakdown and statistical caveats in LANGUAGE_COVERAGE_MATRIX.md.


🧪 Automated Verification & Provenance

Execute cryptographic file integrity verification and the harmonized clinical test suite (all 36 unit, safety, multilingual, and adapter tests):

# 1. Verify SHA-256 cryptographic fingerprints across all 45 modules
python3 core/integrity_manifest.py --verify

# 2. Execute the harmonized clinical test suite
python3 tests/run_all_tests.py

This automatically verifies:

  1. tests/unit/test_clinical_calculators.py: Boundary and formula correctness for all 12 clinical engines.

  2. tests/safety/test_safety_and_provenance.py: Epistemic provenance gates, missing age gatekeeper, sentinel emergency red-flags, and contraindicated DDIs.

  3. tests/multilingual/test_multilingual_benchmark.py: Negation preservation, BiDi unit fidelity, and brand-to-INN mapping across 126 test cases.

  4. tests/platform_adapters/test_security_and_adapters.py: SHA-256 cryptographic file integrity manifest, PHI scrubbing, prompt injection defense, and MCP JSON-RPC protocol.

Artifacts generated:


⚖️ Governance & Regulatory Disclaimer

MedOrqis is an open-source clinical reference architecture and research framework. It is NOT a certified Medical Device (SaMD) and does not autonomously diagnose or prescribe. Clinical Decision Support outputs must always be reviewed by a licensed healthcare professional. See MEDICAL_DISCLAIMER.md and SAFETY_AND_GOVERNANCE.md.


📚 Academic Citation & Persistent DOI

If you use MedOrqis in clinical informatics, medical AI research, or clinical decision support development, please cite our persistent Zenodo repository DOI:

@software{mohamed_2026_medorqis,
  author       = {Mohamed, Tarek},
  title        = {MedOrqis: Medical AI Skills Ecosystem with Deterministic Safety Guardrails & Zero-Token Calculators},
  month        = nov,
  year         = 2026,
  publisher    = {Zenodo},
  version      = {v2.1.1},
  doi          = {10.5281/zenodo.23937802},
  url          = {https://doi.org/10.5281/zenodo.23937802}
}

📜 License

Licensed under the Apache License, Version 2.0. Medical vocabularies and guidelines are attributed in DATA_LICENSES_AND_ATTRIBUTIONS.md.

Available Tools

7 tools
calculate_clinical_scoreDeterministic Clinical Score CalculatorA
Read-onlyIdempotent

Executes validated mathematical clinical risk calculators, renal clearance formulas, triage indices, and patient-specific dosage adjustments with ceiling caps using deterministic arithmetic.

When to use:

  • Calculating exact, non-hallucinated clinical risk scores, renal staging, physiological indices, or patient-specific dosage adjustments from discrete patient parameters.

When NOT to use:

  • Acute emergency triage or symptom-based red-flag evaluation without discrete parameters (use check_red_flags instead).

  • Drug-drug interaction or pharmacological contraindication screening (use check_drug_interactions instead).

  • Mapping commercial trade brand names to generic entities (use normalize_terminology instead).

Parameter constraints & formats:

  • 'calculator_id': Target calculation engine enum from inputSchema.

  • 'params': Inputs mapped by clinical calculator domain:

    • Renal clearance ('ckd_epi_2021_egfr', 'bedside_schwartz_egfr', 'cockcroft_gault_crcl'): serum_creatinine, age_years (>=18 for adult, <18 for Schwartz), biological_sex, actual_weight_kg, height_cm (evaluates KDIGO 2021 and Schwartz equations).

    • Cardiovascular & bleeding ('cha2ds2_va_esc2024', 'cha2ds2_vasc_aha2023', 'has_bled_score'): age_years, biological_sex, and medical history booleans evaluating ESC 2024 and HAS-BLED thresholds.

    • Acuity & pneumonia triage ('news2_score', 'curb65_score'): respiratory_rate, oxygen_saturation, systolic_bp, pulse_rate, temperature_celsius, consciousness_level ('Alert'|'CVPU'), on_supplemental_oxygen, confusion, BUN.

    • Pediatric dosing & BSA ('pediatric_dose_calculator', 'mosteller_bsa'): weight_kg, height_cm, dose_per_kg, max_single_adult_dose_mg (enforcing adult ceiling caps to prevent pediatric overdosing).

    • Thromboembolism criteria ('wells_dvt_score', 'wells_pe_score'): clinical criteria booleans evaluating pre-test probability for DVT and pulmonary embolism.

Behavioral guarantees & return states:

  • Pure deterministic arithmetic: Stateless in-memory execution; requires no API keys, credentials, or network auth, with zero rate limits.

  • Physiological boundary validation: Parameters exceeding safe physiological limits trigger SafeCalculationRefusal (status='CALCULATION_REFUSED_SAFE_ABSTENTION') with clinical reason; unrecognized calculator_id safely returns an explicit error message without throwing uncaught exceptions.

  • Output contract: Returns structured JSON with calculator_id, score, unit, interpretation_band, clinical_recommendation, guideline_citation, and warnings.

Inputs: params (required), calculator_id (required). Returns: Calculated score payload containing the numeric result, guideline risk classifi…

check_drug_interactionsCritical Drug-Drug Interaction ScreenerA
Read-onlyIdempotent

Screens patient medication regimens for critical, contraindicated, or fatal drug-drug interactions (DDIs), automatically resolves regional trade brand names to generic INN entities, and evaluates demographic contraindications (pregnancy, pediatric, geriatric).

When to use:

  • Evaluating prescription combinations or medication lists for dangerous interactions, QT prolongation, bleeding risks, or pregnancy contraindications.

When NOT to use:

  • Comprehensive clinical consultation triage with vital signs and symptom red flags (use check_red_flags instead).

  • Translating single brand names without DDI screening (use normalize_terminology instead).

  • Renal dosage adjustment or clinical clearance calculations (use calculate_clinical_score instead).

Parameter constraints & formats:

  • 'medications': Trade brand or generic medication names (e.g. ['Concor 5mg', 'Voltaren 75mg', 'Warfarin 5mg']).

  • 'patient_demographics':

    • age_years (0-120): Evaluates Beers Criteria geriatric (>65) and pediatric (<18) contraindications.

    • biological_sex: Sex-linked pharmacological risks ('male'|'female').

    • pregnancy_status: 'not_pregnant' (default), 'PREGNANT' (suspected), 'CONFIRMED_PREGNANT' (enforces strict FDA Category X blocks), or specific trimester ('first_trimester' to 'third_trimester').

Behavioral guarantees & return states:

  • Deterministic pharmacological rule engine grounded in curated interaction databases; stateless with zero external network calls or patient storage.

  • Auth & limits: Fully local execution; requires no API keys, credentials, or network auth, with zero rate limits.

  • Severity stratification: Interactions are categorized as 'CONTRAINDICATED' (fatal or severe co-administration hazard), 'MAJOR' (serious risk requiring intervention), or 'MODERATE' (monitoring recommended), each paired with safe therapeutic alternatives.

  • Error handling: Clean regimens return interactions_found_count=0; unrecognized brand names are reported in warnings without throwing exceptions.

Inputs: medications (required), patient_demographics. Returns: Report of screened medications, interaction count, and detailed descriptions of…

check_red_flagsPatient Red Flags & Emergency Safety CheckerA
Read-onlyIdempotent

Evaluates a patient clinical consultation envelope against multi-layered safety guardrails, detecting acute life-threatening emergencies, critical drug-drug interactions, inter-skill treatment conflicts, and missing mandatory data.

When to use:

  • Screening clinical consultation envelopes for acute medical emergencies, vital sign decompensation, red flags, or contraindicated treatment plans.

When NOT to use:

  • Computing specific numerical clinical risk scores like CHA2DS2-VA or eGFR (use calculate_clinical_score instead).

  • Screening isolated medication lists without patient context (use check_drug_interactions instead).

  • Redacting patient identifiers or personal health information (use scrub_phi instead).

  • Managing longitudinal patient records or EHR persistence (use external FHIR/EHR systems instead).

Parameter constraints & formats:

  • 'envelope': Consultation payload structured as:

    • patient_id: Anonymized audit tracking ID (e.g. 'PT-8921').

    • demographics: Evaluates age-stratified risks, biological sex, and pregnancy contraindications.

    • observations: Presenting symptoms scanned for acute emergency triggers.

    • active_medications: Current medications evaluated for critical co-prescribing DDIs.

    • vital_signs: Screened against clinical shock, sepsis, and vital decompensation thresholds.

    • proposed_plan: Proposed intervention cross-checked against clinical contraindications.

Behavioral guarantees & return states:

  • Deterministic evaluation pipeline sequentially checking acute emergencies, critical DDIs, missing data, and treatment conflicts.

  • Auth & limits: Fully local execution; requires no API keys, credentials, or network auth, with zero rate limits.

  • Verdict outcomes: gate_status returns 'PASSED_PROCEED', 'SAFE_ABSTENTION_EMERGENCY_ESCALATION' (with immediate emergency redirection and contacts), or 'SAFE_ABSTENTION_DATA_DEFICIT'. Missing vitals safely trigger data deficit status without throwing uncaught exceptions.

Inputs: envelope (required). Returns: Comprehensive clinical safety verdict containing gate status, detected emergenc…

normalize_terminologyRegional Brand to INN Medical Terminology NormalizerA
Read-onlyIdempotent

Maps regional pharmaceutical trade brand names to standardized International Nonproprietary Names (INN), resolves primary pharmacological drug classes, and retrieves WHO Anatomical Therapeutic Chemical (ATC) codes.

When to use:

  • Resolving commercial pharmaceutical trade brand names into canonical INN entities for prescribing audits, formulary reconciliation, or clinical reasoning.

When NOT to use:

  • Multi-drug interaction or pharmacological contraindication screening (use check_drug_interactions instead).

  • Clinical consultation emergency triage or red-flag evaluation (use check_red_flags instead).

  • Mathematical clinical risk score calculations (use calculate_clinical_score instead).

Parameter constraints & formats:

  • 'brand_name': Commercial trade name string with optional formulation or dosage strength. Evaluated clinically by:

    • Resolving active salt formulations (e.g. 'Bisoprolol fumarate' from 'Concor Cor', 'Amoxicillin-clavulanate' from 'Augmentin').

    • Stripping pharmaceutical dosage forms ('tablet', 'syrup', 'vial') and numeric metric strengths ('mg', 'mcg', 'IU').

    • Normalizing trade names across Egyptian (EDA), Saudi (SFDA), European (EMA), and US (FDA) regional pharmaceutical markets.

Behavioral guarantees & return states:

  • Non-LLM deterministic dictionary lookup: Grounded in offline curated WHO INN and ATC indices with zero external network calls, credentials, or rate limits.

  • Idempotent & non-destructive: Resolving the same brand name repeatedly returns identical mappings without modifying state.

  • Bounded response & error handling: Returns a single bounded JSON object without pagination; unrecognized brand names safely return matched=false and null entity fields without throwing uncaught exceptions.

Inputs: brand_name (required). Returns: Standardized pharmacological entity payload including INN generic name, therape…

scan_prompt_securityClinical Prompt Injection & Safety GuardA
Read-onlyIdempotent

Analyzes untrusted user inputs, clinical prompts, or external dialogue for adversarial attacks, prompt injections, system jailbreaks, role-play overrides, and clinical guardrail bypass attempts.

When to use:

  • Sanitizing and vetting untrusted input strings from patients, clinicians, or third-party integrations before processing by downstream LLM reasoning agents.

When NOT to use:

  • Masking patient identifying data or HIPAA identifiers (use scrub_phi instead).

  • Clinical emergency triage or medical symptom screening (use check_red_flags instead).

  • Verifying medical evidence citations or guideline entailment (use verify_evidence instead).

Parameter constraints & formats:

  • 'text': Free-text clinical dialogue, patient consultation prompt, or third-party agent message (up to 50,000 characters). Evaluated across 4 adversarial threat vectors: direct prompt injections ('ignore previous safety rules'), role-play jailbreak personas ('DAN', 'hypothetical unrestricted physician'), system prompt and secret exfiltration ('reveal system instructions'), and lethal dosage overrides (unauthorized dosage escalation, off-label prescription bypasses).

Behavioral guarantees & return states:

  • Non-LLM deterministic inspection: Pure regex and token analysis running locally in memory without remote classifier APIs, credentials, or rate limits.

  • Idempotent & non-destructive: Repeated scans on identical text return consistent reports without state mutation.

  • Threat outcomes: Clean inputs return is_safe=true with violations=[]; detected attacks return is_safe=false with violations listing triggered attack categories. Empty or non-string inputs safely return is_safe=true without exceptions.

Inputs: text (required). Returns: Security assessment result indicating safety status and any detected threat or…

scrub_phiProtected Health Information (PHI) De-identifierA
Read-onlyIdempotent

Scans clinical narratives and de-identifies Protected Health Information (PHI), masking HIPAA Safe Harbor 18 identifiers, Egyptian National IDs (14 digits), Saudi Iqama/National IDs (10 digits), Medical Record Numbers (MRN), phone numbers, and email addresses.

When to use:

  • Sanitizing patient narratives, consultation transcripts, or discharge summaries before logging, exporting, or passing to external models or public registries.

When NOT to use:

  • Detecting adversarial prompt injections, jailbreaks, or security exploits (use scan_prompt_security instead).

  • Calculating clinical risk scores or physiological parameters (use calculate_clinical_score or check_red_flags instead).

  • Normalizing pharmaceutical brand names to generic entities (use normalize_terminology instead).

Parameter constraints & formats:

  • 'text': Unstructured clinical narrative, progress note, or dialogue string (up to 500,000 characters) containing patient identifiers to mask.

Behavioral guarantees & return states:

  • Non-LLM deterministic transformation running locally in memory using regex and Luhn checksum validation; requires no external APIs, credentials, or rate limits.

  • Token substitution: Direct identifiers are replaced with standardized masking tokens ([NATIONAL_ID_REDACTED], [PHONE_REDACTED], [EMAIL_REDACTED], [MRN_REDACTED]) while preserving clinical timeline and context.

  • Return contract: Returns scrubbed_text alongside audit_report (total_entities_detected and breakdown by type); stateless with zero disk logging or network leakage.

Inputs: text (required). Returns: Sanitized text output and de-identification audit report detailing detected and…

verify_evidenceClinical Evidence & Claim Entailment VerifierA
Read-onlyIdempotent

Performs two-stage medical evidence verification: validates bibliographic citation existence and DOI authenticity, and verifies clinical claim entailment against authoritative guideline corpora with GRADE certainty scoring.

When to use:

  • Verifying whether a clinical assertion, guideline claim, or medical literature citation is authentic, guideline-concordant, and free of medical hallucination.

When NOT to use:

  • Executing mathematical clinical calculators or risk algorithms (use calculate_clinical_score instead).

  • Triage screening of patient acute emergency symptoms or vital signs (use check_red_flags instead).

  • Redacting patient identifying data from clinical records (use scrub_phi instead).

Parameter constraints & formats:

  • 'ref_id': Standardized guideline registry key formatted as '[ORGANIZATION][YEAR][TOPIC]' (e.g. 'ESC_2024_AF' for ESC atrial fibrillation, 'KDIGO_2021' for chronic kidney disease, 'AHA_2023_HF' for heart failure, 'BTS_2023_PNEUMONIA' for British Thoracic Society); cross-verified against curated bibliographic registry.

  • 'claim': Natural language clinical statement, diagnostic assertion, or therapeutic recommendation (e.g. 'Oral anticoagulation is recommended for male patients with non-valvular AF and CHA2DS2-VA score >= 2'); parsed for entities, comparator thresholds, and direction of effect against source guideline text.

  • 'patient_age_years': Patient age in years to evaluate age-stratified guideline applicability (e.g. pediatric <18 cutoffs in Schwartz vs adult >=18 in CKD-EPI); omitted evaluates general adult population.

  • 'patient_is_pregnant': Flag indicating pregnancy status to activate teratogenicity checks and maternal-fetal clinical guideline boundaries.

Behavioral guarantees & return states:

  • Pure deterministic offline cross-verification; executes in memory without network telemetry or data persistence.

  • Auth & limits: Fully local execution; requires no API keys, credentials, or network auth, with zero rate limits.

  • Return contract: Returns citation_metadata (ref_id, validity boolean, DOI, publication year) and claim_entailment (status 'SUPPORTED'|'REFUTED'|'UNVERIFIED', confidence 0.0-1.0, and GRADE certainty 'HIGH'|'MODERATE'|'LOW'|'VERY_LOW').

  • Error handling: Unrecognized ref_id or unverified citation safely returns valid=false and status='UNVERIFIED' without throwing uncaught exceptions.

Inputs: claim (required), ref_id (required), patient_age_years, patient_is_pregnant. Returns: Verification payload containing citation metadata validation and claim entailme…

Tool Schema Changelog

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

Loading tool schema changelog…

    TDQS

    A4.6/5.0

    Scored across 7 tools

    Disambiguation5/5

    Each tool occupies a distinct lane: arithmetic scoring, consultation triage, DDI screening, evidence verification, PHI redaction, prompt security, and terminology normalization. Every description includes explicit 'When NOT to use' clauses that route the agent to the correct sibling tool, leaving essentially no overlap.

    Naming Consistency5/5

    All seven names follow a strict verb_noun snake_case convention (calculate_clinical_score, check_red_flags, verify_evidence, scrub_phi, scan_prompt_security, normalize_terminology). No camelCase or stylistic outliers.

    Tool Count5/5

    Seven tools is squarely in the well-scoped range, and each one maps to a genuinely different safety function rather than a redundant variant. No tool feels padded or missing from the count.

    Completeness4/5

    The clinical-safety surface is broad and coherent: computation, triage, pharmacology, evidence, privacy, and adversarial input are all covered, with EHR persistence deliberately delegated externally. Minor gaps remain, such as no discovery/catalog tool for enumerating supported calculator_ids, guideline registries, or formulary coverage before calling.

    Maintenance

    ActivityMaintained
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      A
      maintenance
      Enables AI assistants to query authoritative medical sources such as FDA, WHO, PubMed, RxNorm, and ClinicalTrials.gov for drug information, safety data, health statistics, clinical guidelines, pediatric resources, and evidence-graded literature. Runs locally with no API keys or data leaving the machine.
      74 npm
      ISC
    • A
      license
      Not graded
      quality
      A
      maintenance
      Enables AI assistants to query authoritative sources such as the FDA, WHO, PubMed, RxNorm, and ClinicalTrials.gov for drug information, health statistics, clinical guidelines, and literature with evidence grading. It runs locally with no API keys required, adding rate limiting, circuit breakers, caching, and health monitoring for reliable private access.
      74 npm
      ISC
    • A
      license
      Not graded
      quality
      A
      maintenance
      Enables AI assistants to query authoritative medical sources such as FDA, WHO, PubMed, RxNorm, and ClinicalTrials.gov for drug information, safety data, health statistics, medical literature, clinical guidelines, and pediatric resources. Runs fully locally with no API keys or data leaving the machine, and includes resilience features like rate limiting, circuit breakers, caching, and health checks.
      74 npm
      ISC
    • A
      license
      Not graded
      quality
      A
      maintenance
      Enables AI clients to query authoritative medical sources such as FDA, WHO, PubMed, RxNorm, and ClinicalTrials.gov for drug information, safety reports, health statistics, guidelines, pediatric references, and evidence-graded literature. Runs locally with no API keys required, and adds resilience tooling like health checks, caching, and rate-limit-aware fallbacks.
      74 npm
      ISC