SCOPE Method - Nano PRD Template
The SCOPE Method for AI agent specs is a specific framework for writing lean feature specs — not a generic project scoping tool. Based on the research from the previous turn, here’s the full template
[Feature/Epic Name]
S — Situation (Why)
What's broken/missing? Include user impact/business signal (1-3 sentences).
Ex: Users see generic "Error" on card failure, abandoning checkout. Losing ~12% of attempts to recoverable errors.
C — Constraints (No-Gos & NFRs)
Rules the implementation MUST NOT break (compliance, UX, tech debt).
- [Rule] — [Reason]
Ex: No multi-step email flows (1 max) | No raw card data in logs (PCI) | Preserve user inputs on retry (UX)
O — Outcome (Acceptance Criteria)
Testable Given/When/Then statements defining "done".
- Given [context], when [action], then [expected result]
Ex: [ ] Given an expired card, when submitted, then show inline error "Card expired" (no modal).
P — Patterns (Tech Context)
Stack, components, APIs, or conventions to follow (prevents AI hallucinations).
- Stack/API: [e.g., Next.js 14, Stripe SDK v12, Stripe error codes]
- Reuse: [e.g.,
<ErrorBanner>at/components/ui/] - DB: [e.g., add
recovery_sentboolean tocheckout_sessions]
E — Edge Cases
Failure states, empty states, race conditions, and boundary conditions.
- What if [X fails/duplicate trigger]? → [Fallback/Guard]
Ex: Job fires twice? →
recovery_sentprevents double-send | User pays after email queued? → Cancel job.
Dev Notes & Success Metric
- Dev Notes: [Implementation decisions/approaches chosen by devs]
- Success Metric: [One signal proving production success] (Ex: Abandon rate drops <5% in 7 days)
Created on 5/14/2026