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). 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 for people whose travel insurance PDF is 47 pages and their flight leaves in 3 hours."_

**Summary:** A conversational AI agent that triage-diagnoses your emergency, then delivers a ranked, time-sensitive action plan with real contacts, real addresses, and real document checklists — specific to your country, situation, and time of day.

## Agent-readiness score

Overall: **54/100** (band D)

| Dimension | Score | Why |
|---|---|---|
| Memory required | 20/25 | 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 | 9/25 | Wide policy surface — full red-team pass, content filter, and human-in-loop required. |
| Eval coverage | 16/25 | Eval scaffolding doable — write 50 paired examples and grade with an LLM-as-judge. |

> Build only if you have a moat. PanicPal 911's readiness gap is real work.

## 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)

## 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 Sitata?", the agent gives a concrete differentiator, not a marketing line.
- When asked about Google'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: `Next.js`, `Claude API (Anthropic)`, `Supabase`, `OpenStreetMap Nominatim API`, `Twilio (SMS fallback for low-connectivity emergencies)`
- Solo build estimate: 3-4 months for a solid MVP covering 15 core scenarios

## Kill prediction

Google could obsolete this in 12-18 months. Google Maps already knows where you are, Gemini already speaks 40 languages, and Google Pay already knows your cards. One 'Emergency Mode' feature in Google Maps wipes out the entire category before you finish your Series A pitch deck.

**Survival strategy:** Go deep B2B into corporate travel management (Concur, TravelPerk integrations) before Google notices. Own the enterprise liability-indemnified contract layer Google won't touch.

## Hand-off

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