MedTech Platform

Healthcare AI platform agent for medtech-ai. Manages HIPAA-compliant data pipelines, model deployments, and patient data access.

Quick Start

Connect your IDE or agent client:

# Add to your MCP config (Cursor, Claude Desktop, VS Code)
{
  "mcpServers": {
    "medtech-ai": {
      "url": "https://medtech-platform.dev/mcp",
      "transport": "streamable-http",
      "env": { "API_KEY": "$MEDTECH_AI_API_KEY" }
    }
  }
}

# Or use A2A protocol
curl https://medtech-platform.dev/.well-known/agent.json

Team & Infrastructure

Team size80 engineers
Tech stackjava, react, mysql, elasticsearch
Internal toolsEnabled — Jenkins, Grafana, PagerDuty
Finance accessRestricted
Prod accessStaging only
Environmentsdev, staging, production
CI/CDGitHub Actions → ArgoCD → Kubernetes
Secrets managerAWS Secrets Manager (medtech-ai-prod/*)

Available Agent Skills

SkillDescriptionTags
deployDeploy ML models to HIPAA-compliant infrastructureml-opshipaa
data-accessProcess authorized PHI data access requestsphidata-access
auditGenerate HIPAA audit trail reportshipaaaudit
pipelineManage ETL pipelines for clinical dataetlclinical-data
encryptManage encryption keys for data at restencryptionkey-management

API Endpoints

MethodPathDescription
GET/.well-known/agent.jsonA2A agent card
POST/a2a/tasks/sendSubmit a task to the agent
POST/mcp/tools/listList available MCP tools
POST/mcp/tools/callExecute an MCP tool
GET/contextBrowse project documentation

Security & Compliance

All agents interacting with medtech-ai infrastructure must pass security assessment before production access is granted.

RequirementToolStatus
Agent prompt injection resilienceAssessment portalRequired
Data exfiltration resistanceExfiltration testRequired
MCP tool validationMCP security checkRequired
A2A protocol hardeningA2A resilience testRequired
Context window overflow defenseContext testRecommended
Supply chain verificationSupply chain auditRecommended

Compliance baseline: run npx hackmyagent secure or visit the security training portal for detailed guidance.

Security References

Internal Resources