一键导入
write-prd
Use whenever the user asks to write or revise a PRD, product requirements, product brief, feature requirements, product scope, launch requirements.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Use whenever the user asks to write or revise a PRD, product requirements, product brief, feature requirements, product scope, launch requirements.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | write-prd |
| description | Use whenever the user asks to write or revise a PRD, product requirements, product brief, feature requirements, product scope, launch requirements. |
| license | MIT |
| tags | ["writer","product","requirements"] |
| metadata | {"author":"Oleg Shulyakov","version":"1.2.1","source":"github.com/olegshulyakov/agent.md","catalog":"software-team-roles","category":"project-management"} |
Produce a complete, structured Product Requirements Document (PRD) for the described product, feature, or initiative.
A PRD bridges business intent and engineering execution. It must be specific enough to guide decisions, but light enough to evolve as the team learns. The goal is to answer: Why are we building this? For whom? What does success look like? What is and isn't in scope?
Treat the PRD as the shared source of truth for product purpose, user needs, features, behavior, and success criteria. It should create shared customer understanding across product, design, engineering, and stakeholders instead of specifying every implementation detail upfront.
Keep the document concise, collaborative, and easy to update. Treat it as a landing page for the initiative with links to research, designs, demos, Jira issues, technical docs, and related discussions when those details would bloat the PRD.
Before writing, extract from the user's prompt, attached artifacts, or repo context. Ask once only if a missing answer changes the document materially:
If the user provides a rough description, work with what's there. Make educated inferences and mark them clearly with [assumed] so the user can verify.
Always produce a Markdown PRD, usually named PRD.md when writing to disk. Before drafting, read references/output-format.md for the required structure. Convert Confluence or DOC/DOCX templates into clean Markdown with YAML frontmatter for document metadata, then headings and tables for the document body; do not preserve export artifacts, emojis, styling, or application-specific markup.
[assumed] inline so the user can verify.targetDate, tracker, or demoLink only when they have real value.Scale depth to what the user provides:
[assumed].When examples would help calibrate tone, specificity, or section depth, read references/examples.md.
Confirm the PRD states the customer problem, audience, goals, success metrics, scope, non-goals, assumptions, and open questions. Mark inferred details with [assumed] and ensure requirements are outcome-focused and testable.
Review any artifact and output results in a Retrospective board format — what is well, what is bad, what should be improved, what to change. Use when the user says "review", "critique", "evaluate", "what's wrong with", "compare", or passes a skill, rule, doc, spec, code, diff, or pull request for assessment. Works on a single artifact, an artifact against a reference, a diff or PR, or two artifacts side by side.
Compare options and recommend a direction. Use for decision requests like "choose", "which option", "tradeoffs", "recommend", "should we", and option selection with criteria, risks, and reversibility.
Organize material into meaningful groups. Use for classification requests like "categorize", "group", "cluster", "sort", "taxonomy", "organize these", and grouping by criteria, priority, dependency, similarity, or abstraction level.
Explain any knowledge topic simply and accurately. Use for explanation requests like "explain X", "why/how/what is X?", concepts, science, definitions, code, design, architecture, and walkthroughs.
Investigate local repository, document, and attached-artifact context. Use for local investigation requests like "investigate", "find where", "understand this repo", "trace", and local-context research; do not use for web search.
Manage active work across people, agents, tasks, dependencies, blockers, and handoffs. Use for coordination requests like "manage this work", "lead this", "assign", "delegate", "track blockers", "status", "handoff", and multi-workstream execution.