Application Test Suite

Run various tests to verify application functionalities and AI model responses.

Core AI Flow

Generate Advice (Basic)

Idle

Tests the generateAdvice flow with minimal user concerns.

Core AI Flow

Generate Clarifying Questions

Idle

Tests the generateClarifyingQuestions flow for initial health concerns.

Core AI Flow

Generate Simulated Physical Exam

Idle

Tests generating physical exam data from EMR summary.

Core AI Flow

Generate Patient Summary

Idle

Tests the generatePatientSummary flow with mock physician selections.

Core AI Flow

Medline Integration

Fetch MedlinePlus Topic

Idle

Tests fetching data from MedlinePlus for a common topic.

Medline Integration

RAG Pipeline

RAG: Convert to Question

Idle

Tests the query to natural language question conversion step of RAG.

RAG Pipeline

RAG: Answer with Context

Idle

Tests the core RAG answering mechanism with mock context.

RAG Pipeline

RAG: Refine Follow-up Question

Idle

Tests refining a follow-up question based on an original query.

RAG Pipeline

Differential Diagnosis Logic

De-duplicate Medical Conditions

Idle

Tests de-duplication and standardization of medical condition names.

Differential Diagnosis Logic

Enrich Manually Added Diagnosis

Idle

Tests AI enrichment for a manually added diagnosis within patient context.

Differential Diagnosis Logic

Diagnose Follow-up Question

Idle

Tests AI response to a physician follow-up on a specific diagnosis.

Differential Diagnosis Logic

Generate Specific Clarifying Questions

Idle

Tests AI generation of targeted clarifying questions for a diagnosis.

Differential Diagnosis Logic

Performance

Performance: Generate Advice

Idle

Measures execution time for the generateAdvice flow.

Performance

Performance: Full RAG Pipeline (Simplified)

Idle

Measures execution time for a simplified RAG query.

Performance

Robustness Check

Robustness: Out-of-Domain Query (Advice)

Idle

Tests generateAdvice with a non-medical query. Expect polite refusal or generic non-answer.

Robustness Check

End-to-End

E2E: Advice Form (Initial AI Clarifying Questions)

Idle

Simulates the flow: concern -> initial AI clarifying Qs -> final advice based on those answers.

End-to-End

E2E: Advice Refinement (Physician Clarifying Questions)

Idle

Simulates advice refinement after physician asks specific clarifying questions and gets answers.

End-to-End