How These Examples Work

Each example demonstrates an AI agent reading from and writing to a Cascade Protocol Pod. All processing runs locally on your machine — no health data leaves your device.

💾

Reference Patient Pod

30 days of realistic data for a fictional patient, included in this repository.

🔧

Two Access Methods

Use the Cascade MCP server with Claude Desktop, or invoke the Cascade CLI directly with Claude Code.

📄

Prompt + Expected Output

Each example includes the exact prompt and a reference answer to evaluate the agent's response.

← Back to Documentation Security & Agent Consent