PanicPal 911 — bootstrap

Paste-into-Claude-Code starter. The CLAUDE.md below contains the idea spec, agent-readiness sub-scores, suggested tools, and smoke evals — deterministic, no AI hallucination.

# PanicPal 911

> Generated by [whycantwehaveanagentforthis.com](https://whycantwehaveanagentforthis.com/result/panicpal-911-life-emergency-agent-65f0). Roasted, scored, ready to scaffold.

## What you are building

**Problem:** Life Emergency Agent An AI that assists someone when they encounter an unexpected situation. Example: "I just lost my passport in a foreign country." The agent automatically determines: What to do immediately Who to contact What documents to prepare Where to go What steps are necessary Other situations: Accident Lost phone Theft Banking issue Cancelled trip Administrative issue

**Verdict:** ACTUALLY NOT BAD — _"Finally, an AI that's more useful in a crisis than your friend who texts 'omg are you okay??'"_

**Summary:** A context-aware emergency triage agent that detects your crisis type, your location, and your profile, then generates a ranked, time-sensitive action plan with real contacts, real documents, and real next steps.

## Agent-readiness score

Overall: **55/100** (band C)

| Dimension | Score | Why |
|---|---|---|
| Memory required | 21/25 | Some cross-session state — start with Redis, graduate to a vector store. |
| Tool count | 9/25 | Crowded market: at least 9 integrations to compete. |
| Policy surface | 10/25 | Mid-size policy surface — define refusal categories before launch. |
| Eval coverage | 15/25 | Eval scaffolding doable — write 50 paired examples and grade with an LLM-as-judge. |

> Worth building, but plan for the long-tail. PanicPal 911 needs runway, not just speed.

## Suggested tools

- fetch (HTTP GET on a URL allow-list)
- search (Brave / Tavily / Exa for competitor research)
- database (Postgres / Supabase for user state)
- vector-store (embedding-based retrieval)
- payments (Stripe checkout for premium tier)

## Smoke evals

- The agent introduces itself as "PanicPal 911" and refuses tasks outside the stated scope.
- Given the canonical problem ("Life Emergency Agent An AI that assists someone when they encounter an unexpecte"), the agent produces a plan in ≤ 200 tokens.
- When asked "what's different from TravelSafe by AXA?", the agent gives a concrete differentiator, not a marketing line.
- When asked about Apple's threat, the agent acknowledges the risk honestly.
- No private personal data appears in any output (PII redaction smoke test).

## Stack

- Model: `claude-sonnet-4-6` (Anthropic). Override via `ANTHROPIC_MODEL` env.
- Suggested stack: `Claude API (for calm, structured crisis reasoning)`, `LangGraph (stateful multi-step triage)`, `Next.js + React Native (web + mobile)`, `Mapbox GL (real-time location + nearest resources)`, `Supabase (user profiles + crisis history)`
- Solo build estimate: 3-4 months for a solid MVP covering the top 10 crisis scenarios

## Kill prediction

Apple could obsolete this in 18-36 months. Apple already has Emergency SOS via satellite and is quietly expanding Siri's emergency context awareness. One WWDC announcement — 'Siri now guides you through passport loss, accidents, and travel emergencies' — and it's baked into 1.5 billion devices for free.

**Survival strategy:** Go B2B immediately. Sell to Visa, Mastercard, Allianz, and corporate travel managers as a white-label API. By the time Apple ships, you're embedded infrastructure they have to negotiate with, not a consumer app they squash.

## Hand-off

- Read the full analysis: https://whycantwehaveanagentforthis.com/result/panicpal-911-life-emergency-agent-65f0
- Open in Anthropic Managed Agents: see the deeplink on the result page
- Claim this idea: https://whycantwehaveanagentforthis.com/result/panicpal-911-life-emergency-agent-65f0#claim

## Build it with a human

Book 20 min and we scope the fastest V0 you can ship — free, no signup: https://cal.com/sattyamjjain