◂ Latent · 潜
Application demo · RAD / AI-assisted full-stack mini app
应用 Demo · RAD / AI 协助全栈 Mini App

Generating an expense-approval app from one sentence

从一句话生成报销审批 App

Vibe coding made visible: the primary plane isn't "writing the 7th file" — it's the agent's understanding of "what to build" — the requirement, the architecture decision (hung on a re-runnable smoke), the next plan, the undecided notification-channel option, and honest risks like "attachment storage is a production risk." The same CognitiveState drives this interface and is exactly what the latent-design skill teaches a coding agent to emit.

vibe coding 的现身说法:主舞台不是“正在写第 7 个文件”,而是 agent 对“要构建什么”的理解——需求 requirement、架构 decision(挂在可重跑的 smoke 上)、下一步 plan、未定的通知渠道 option、以及“附件存储是生产风险”这样的诚实 risk。同一份 CognitiveState 既驱动这个界面,也正是 latent-design skill 教 coding agent 去 emit 的东西。

AI-assisted full-stack mini app · expense approvals
Generate a working expense-approval app from a one-line requirement (frontend + API + storage + auth)
Understand the askv0 generated
Your question
Wanted a small expense-approval tool and gave a one-line requirement. Get a working v0 and know what to decide next
My read
still forming — don't trust me yet…
Why I think so
Requirement as understood: three roles (employee/manager/finance) + attachments + state transitions + auth fairly sure
Because roles and flow unpacked from the one-line requirement.
how I confirmed this ▾
I considered but ruled out 1 idea ▾