to-prd
The to-prd skill synthesizes current conversation context and existing codebase understanding into a structured Product Requirements Document without re-interviewing the user. It generates implementation-ready PRDs using domain-specific language to articulate problems, goals, user stories, implementation decisions, testing boundaries, and explicit out-of-scope items, while referencing existing Architecture Decision Records to avoid rehashing prior decisions. Use this skill when users explicitly request a PRD or need to formalize existing ideas into structured requirements for issue tracking systems.
git clone --depth 1 https://github.com/Haohao-end/openagent /tmp/to-prd && cp -r /tmp/to-prd/api/internal/core/skills/catalog/to-prd ~/.claude/skills/to-prdskill.md
# 转成 PRD 这是一个把当前上下文整理成 PRD 的技能。 ## 适用场景 - 用户明确要 PRD - 用户希望我把现有想法变成结构化需求文档 - 需要把方案沉淀到 issue tracker 里，供后续实现使用 ## 工作方式 1. 基于当前对话和已有代码库理解直接综合，不重新采访用户。 2. 用项目领域语言写清楚问题、目标、用户故事、实现决策、测试决策和不做什么。 3. 如果已有相关 ADR，沿用既有决策，不要重提旧争论。 4. 输出要能直接交给实现者。 ## 约束 - 不要写成泛泛的愿景稿。 - 不要塞进具体文件路径或临时代码片段。 - 不要遗漏测试边界。
Use when the user wants a less default, more original solution for code, design, writing, naming, architecture, debugging, or strategy, especially when progress is stuck.