SpendSherlock 5000: The Reckoning — 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.

# SpendSherlock 5000: The Reckoning

> Generated by [whycantwehaveanagentforthis.com](https://whycantwehaveanagentforthis.com/result/spendsherlock-5000-the-reckoning-spendsherlock-5000-battle). Roasted, scored, ready to scaffold.

## What you are building

**Problem:** SpendSherlock 5000: battle continuation

**Verdict:** ACTUALLY NOT BAD — _"You named your agent better than most YC founders name their entire company. Respect."_

**Summary:** SpendSherlock 5000 is an AI agent that continuously monitors your transactions, detects suspicious patterns, identifies waste and subscriptions you forgot about, and delivers brutally honest spending narratives — like a detective who moonlights as your disappointed accountant.

## Agent-readiness score

Overall: **56/100** (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 | 9/25 | Wide policy surface — full red-team pass, content filter, and human-in-loop required. |
| Eval coverage | 17/25 | Eval scaffolding doable — write 50 paired examples and grade with an LLM-as-judge. |

> Worth building, but plan for the long-tail. SpendSherlock 5000: The Reckoning 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 "SpendSherlock 5000: The Reckoning" and refuses tasks outside the stated scope.
- Given the canonical problem ("SpendSherlock 5000: battle continuation"), the agent produces a plan in ≤ 200 tokens.
- When asked "what's different from Copilot Money?", the agent gives a concrete differentiator, not a marketing line.
- When asked about Intuit'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`, `Plaid API`, `Claude API (for the detective narrative engine)`, `Supabase`, `Vercel`
- Solo build estimate: 4-6 months to a shippable v1 that doesn't embarrass you at a dinner party

## Kill prediction

Intuit could obsolete this in 18-24 months. They killed Mint, they'll feel guilty, they'll build 'Mint 2.0 with AI' inside TurboTax, spend $200M on it, make it worse than the original, and somehow still capture 40% of the market purely on brand recognition

**Survival strategy:** Own the personality and the community — Intuit cannot do 'fun' and has never successfully built a cult following. If SpendSherlock becomes the brand users quote to their friends, no enterprise clone can replicate that.

## Hand-off

- Read the full analysis: https://whycantwehaveanagentforthis.com/result/spendsherlock-5000-the-reckoning-spendsherlock-5000-battle
- Open in Anthropic Managed Agents: see the deeplink on the result page
- Claim this idea: https://whycantwehaveanagentforthis.com/result/spendsherlock-5000-the-reckoning-spendsherlock-5000-battle#claim