Patient Visit Advocate source-intake idea-001
internal prototype · canonical JSON + Dreamborn Forge HTML
internal generated
source-intake · supabase_json

Patient Visit Advocate source-intake idea-001

source-intake artifact · for Patient Visit Advocate · phase idea-001 · status approved

Planning Surface

Use this to decide what happens next.

Status

approved

Phase

idea-001

Agent Handoff
Start Here

source-intake artifact · for Patient Visit Advocate · phase idea-001 · status approved

Completion Evidence

No explicit evidence field yet. Require tests, screenshots, linked PRs, or reviewed outputs before marking complete.

Artifact Shape
  • meta: source: string, created_at: string, product_stage: string
  • source type: product concept brief
  • anti patterns: 5 items
  • schema version: 1.0
  • raw source summary: AI-first product concept for a patient appointment advocate. Core framing: patients enter doctor visits nervous, unprepared, forgetful, intimidated by time pressure, and unsure what matters. The job is not to give generic medical information; it is to help them show up prepared, think clearly in the moment, and leave with confidence. Avoid the default bolt-on-AI app pattern of symptom checker, static question lists, generic top-10 doctor questions, or chatbot wrapper. Build a thinking system that behaves like a sharp, calm advocate. V1 should focus on conversational intake, a clean one-page doctor visit brief, smart dynamic question generation, and simple post-visit summary. Defer real-time in-room mode until the preparation loop works. Long-term moat is structured longitudinal memory: past visits, symptoms, medications, patterns, and context carried forward responsibly. Under the hood, use orchestrated agents: intake extraction, guardrailed clinical reasoning, question generation, medical/plain-English translation, and structured memory. Hard constraints: not diagnosing, not replacing clinicians, not giving treatment plans, conservative language, escalation guidance for serious symptoms, and explicit uncertainty.
  • extracted principles: 5 items
Structured Payload

Machine-readable source fields

meta
source

atlas-codex intake from user-provided product brief

created at

2026-05-05T03:23:35.513Z

product stage

new project intake

source type

product concept brief

anti patterns
  • Generic symptom checker
  • Static question library
  • Top-10 questions content app
  • Chatbot wrapper with no structured memory or planning system
  • Overconfident medical advice generator
schema version

1.0

extracted principles
  • The real problem is patient cognition under stress, not lack of generic information.
  • The product should organize messy human input into clinical-adjacent clarity without diagnosing.
  • The first value moment is walking into the appointment prepared with a clean brief and smart questions.
  • Realtime appointment mode is compelling but should wait until preparation and debrief work reliably.
  • Longitudinal structured memory is the durable product moat.