Class A (IEC 62304 / MDR) Tier 1 v2.1.0

health-cybersecurity-privacy

Clinical Governance, Orchestration & Privacy

Test in Playground GitHub

Intended Clinical Purpose

Screens clinical inputs for PHI and adversarial injections before transmission or public storage.

Prohibited Clinical Uses & Gating Boundary

Transmitting identified patient data to public services

Mandatory Input Parameters

clinical_text

Deterministic Dependencies

core.phi_deidentifier core.prompt_security_guard

Clinical Review Status: REFERENCE_SEED_REQUIRES_CLINICAL_SIGN_OFF

Claude Code & MCP Configuration

Add to Claude Code via MCP stdio:

$ claude mcp add health-cybersecurity-privacy npx medorqis --skill=health-cybersecurity-privacy

Skills CLI Execution

Dispatch directly to agent workspace:

$ npx skills run seotarek/MedOrqis/health-cybersecurity-privacy

SKILL.md & Manifest Specification YAML Frontmatter

---
name: health-cybersecurity-privacy
description: >-
  Health cybersecurity and patient privacy guardian. De-identifies Protected Health Information (HIPAA Safe Harbor, GDPR, Egyptian PDPL, Saudi PDPL) and blocks indirect prompt injection attacks.
allowed-tools: bash
license: Apache-2.0
compatibility: Linux, macOS, Windows (Python 3.10+, MCP stdio, Claude Code, Gemini CLI, OpenAI)
metadata:
  ve...