AgentSkillsCN

Problem-First Workflow Analysis

问题导向型工作流分析

SKILL.md
--- frontmatter
name: "Problem-First Workflow Analysis"
name_zh: "Problem-First Workflow Analysis（问题优先工作流分析）"
category: "strategy-planning"
source: "Lenny's Podcast"
guest: "Keith Coleman & Jay Baxter"

Problem-First Workflow Analysis

Problem-First Workflow Analysis（问题优先工作流分析）

概述 / Overview

一套严谨的研发全生命周期，始于对根本问题的定义，并强制要求在规模化之前，每个阶段都需通过明确的概念验证（POC）。该方法拒绝盲从行业惯例（如人工核查），除非其能通过问题分析的验证。

来源 / Source

  • •嘉宾: Keith Coleman & Jay Baxter
  • •职位: Product Lead & Founding ML Engineer
  • •公司: Community Notes (X)

核心步骤 / Core Steps

  1. •Define Core Problem
  2. •Hypothesis & Mockup Validation
  3. •Closed Pilot (MTurk/Internal)
  4. •Open Public Pilot
  5. •Scale & Expand

核心原则 / Core Principles

  • •Problem Obsession: Focus entirely on solving the core issue (e.g., info quality) rather than business metrics.
  • •Stepwise Proof: The product must prove its value hypothesis at each stage (Mockup -> Pilot -> Scale).
  • •First-Principles Design: Reject existing solutions (Trust & Safety teams) if they don't solve the scale/trust problem.
  • •Radical Transparency: Open-sourcing code and data to build trust rather than relying on brand authority.

适用场景 / When to Use

在低信任、大规模的高风险环境中，从 0 到 1 打造产品。

常见错误 / Common Mistakes

核心机制未验证即开始规模化；依赖用户无法验证的“黑盒”逻辑。

实战案例 / Real-World Example

团队首先利用静态原型（Static Mockups）测试 Community Notes 以验证需求，随后通过 MTurk 任务验证可行性，最后才着手构建线上试点版本（Live Pilot）。

金句 / Quote

"We were very disciplined... about having the product prove itself at every given point."