菜单
Generate a structured PRD from a client brief, written as issues/prd.md. Use to turn a client request or rough idea into a PRD.
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-a-prd |
| description | Generate a structured PRD from a client brief, written as issues/prd.md. Use to turn a client request or rough idea into a PRD. |
| allowed-tools | Read, Bash, Grep, Glob, Write |
This skill will be invoked when the user wants to create a PRD. You may skip steps if you don't consider them necessary.
Ask the user for a long, detailed description of the problem they want to solve and any potential ideas for solutions.
Explore the repo to verify their assertions and understand the current state of the codebase.
Interview the user relentlessly about every aspect of this plan until you reach a shared understanding. Walk down each branch of the design tree, resolving dependencies between decisions one-by-one.
Sketch out the major modules you will need to build or modify to complete the implementation. Actively look for opportunities to extract deep modules that can be tested in isolation.
A deep module (as opposed to a shallow module) is one which encapsulates a lot of functionality in a simple, testable interface which rarely changes.
Check with the user that these modules match their expectations. Check with the user which modules they want tests written for.
issues/prd.md. Create the issues/ directory if it doesn't exist. Do NOT submit a GitHub issue or call any external service.The problem that the user is facing, from the user's perspective.
The solution to the problem, from the user's perspective.
A LONG, numbered list of user stories. Each user story should be in the format of:
This list of user stories should be extremely extensive and cover all aspects of the feature.
A list of implementation decisions that were made. This can include:
Do NOT include specific file paths or code snippets. They may end up being outdated very quickly.
A list of testing decisions that were made. Include:
A description of the things that are out of scope for this PRD.
Any further notes about the feature.
Write, refactor, restructure, edit, or audit technical documentation using the Diátaxis framework (tutorials, how-to guides, reference, explanation). Produces clear, concise, scannable docs and fights the default tendency toward bloated, over-explained AI prose. Use this skill for ANY documentation work — README and CONTRIBUTING files, API/reference docs, runbooks, guides, getting-started pages, conceptual write-ups, docs-site content, or doc reviews — even when the user just says "document this," "write docs for X," "clean up these docs," "explain how this works in the docs," or "what's missing from our documentation." If the deliverable is documentation, reach for this skill rather than writing prose freehand.
Retroactively create Linear tickets and PRs from untracked commits. Use for commits without tickets, "founder mode", or formalizing experimental work.
Create well-structured git commits from analyzed changes. Use when the user asks to commit, save work, or create a commit.
Find architectural friction and propose deep-module refactors as RFC issues. Use to improve architecture, find refactoring opportunities, or deepen shallow modules.
Initialize the AI Engineering Harness (AGENTS.md + thoughts/ structure). Use on /init_harness or when setting up the harness in a repo.
Interview the user relentlessly to stress-test a plan or design. Use to validate design decisions, "drill in", or "ask me questions" before implementation.