agentleFS
dawn840705dawn840705/claude-code-studios/skills/create-prd/SKILL.md

create-prd

Skill1 starsChanged 3 months agoMITView on GitHub

Guided, section-by-section PRD authoring for a single product feature. Gathers context from the product concept and existing PRDs, walks through each required section collaboratively, and writes to product/prd/prd-<feature>.md. Product-track counterpart of /design-system.

44 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

Used with

Projects with it also use prioritize-features, brainstorm-okrs and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • skytiger6724Skill

    prioritize-features

    Prioritize a backlog of feature ideas based on impact, effort, risk, and strategic alignment with top 5 recommendations. Use when prioritizing a feature backlog, making scope decisions, or ranking product ideas.

    Also in 4 projects that use this

  • skytiger6724Skill

    brainstorm-okrs

    Brainstorm team-level OKRs aligned with company objectives — qualitative objectives with measurable key results. Use when setting quarterly OKRs, aligning team goals with company strategy, drafting objectives, or learning how to write effective OKRs.

    Also in 4 projects that use this

  • miovragSkill

    strategy-red-team

    Red-team a PRD, roadmap, or strategy by attacking its load-bearing assumptions before reality does. Steelmans then attacks each claim, ranks failure modes by impact × likelihood × cheapness-to-test, and returns the cheapest test and kill criteria for each. Use when stress-testing a plan, pressure-testing a strategy, challenging assumptions, or preparing a doc for executive review.

    Also in 3 projects that use this

  • launchapp-devSkill

    interview-script

    Create a structured user interview script with warm-up, core exploration, and wrap-up sections. Use when preparing for user research interviews to ensure consistent, insightful conversations.

    Also in 4 projects that use this

  • About-IntelligenceSkill

    beachhead-segment

    Identify the first beachhead market segment for a product launch.

    Also in 4 projects that use this

  • nandanosqlSkill

    customer-journey-map

    Map the complete customer experience across all touchpoints using the NNGroup framework. Identifies pain points, emotions, and opportunities at every stage of the user journey.

    Also in 5 projects that use this

  • +5 more from skytiger6724

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. 0 are common, 9 are distinctive, and 5 are very specific to this project.

  • 0

    Common

    Most of the 42 edited versions we read keep these.

  • 9

    Distinctive

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When this skill is invoked:

    line 9
  2. Very specific:

    If --review [full|lean|solo] was passed → use that

    line 14
  3. Very specific:

    Check if product/prd/product-concept.md exists.

    line 20
  4. Very specific:

    If it exists: read its scope/feature list, find the highest-priority feature without a PRD (product/prd/prd-<feature>.md does not exist), …

    line 21
  5. Distinctive:

    If no product concept exists, fail with:

    line 26
  6. Very specific:

    If product/prd/prd-<feature>.md already exists, do NOT overwrite.

    line 34
  7. Distinctive:

    Read all relevant context before asking the user anything:

    line 42
  8. Distinctive:

    Do not fabricate business facts (pricing, launch dates, user counts) — ask.

    line 61
  9. Distinctive:

    After drafting all sections, show the user a compact summary of what will be written and ask for approval before writing the file:

    line 80
  10. Distinctive:

    Only write after explicit approval.

    line 85
  11. Distinctive:

    All 8 sections present, none placeholder-only

    line 100
  12. Distinctive:

    Every success metric has a number and a measurement source

    line 101
  13. Distinctive:

    Every acceptance criterion maps to an FR

    line 102
  14. Very specific:

    All MVP features covered? → /create-architecture (Architecture phase)

    line 116

“Common” is measured: kept by at least 6 in 10 of the 42 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

When this skill is invoked:

1. Parse Arguments & Validate

Resolve the review mode (once, store for all gate spawns this run):

  1. If --review [full|lean|solo] was passed → use that
  2. Else read production/review-mode.txt → use that value
  3. Else → default to lean

A feature name is required. If missing:

  1. Check if product/prd/product-concept.md exists.
  2. If it exists: read its scope/feature list, find the highest-priority feature without a PRD (product/prd/prd-<feature>.md does not exist), and use AskUserQuestion:
    • Prompt: "The next feature without a PRD is [feature-name]. Author its PRD?"
    • Options: [A] Yes — write the PRD for [feature-name] / [B] Pick a different feature / [C] Stop here
  3. If no product concept exists, fail with:

    "Usage: /create-prd <feature-name> — e.g., /create-prd onboarding No product concept found. Run /product-concept first — it writes product/prd/product-concept.md, which this skill reads."

Normalize the feature name to kebab-case for the filename (e.g., "user onboarding" becomes prd-user-onboarding.md).

If product/prd/prd-<feature>.md already exists, do NOT overwrite. Offer retrofit: identify missing/placeholder sections and fill only those (mirror /design-system retrofit behavior — never touch existing section content).


2. Gather Context (Read Phase)

Read all relevant context before asking the user anything:

  • Product concept: product/prd/product-concept.md — problem, users, value proposition, scope tiers. Fail if missing (see above).
  • Stack pin: .claude/docs/technical-preferences.md — platform, framework, constraints. If missing, warn: stack-unaware PRDs breed infeasible requirements.
  • Existing PRDs: Glob product/prd/prd-*.md — read titles + dependency sections to cross-reference. A new PRD must not contradict or duplicate an existing one.
  • Accessibility tier: design/accessibility-requirements.md if present — requirements below must respect the committed tier.
  • UX specs: Glob design/ux/*.md for screens this feature touches.

3. Author Section by Section (Write Phase)

Walk the user through each required section in order, one at a time: propose a draft from the context you read, ask for corrections, then move on. Do not fabricate business facts (pricing, launch dates, user counts) — ask.

Required sections (all 8 must exist for the PRD to pass review):

  1. Overview & Problem — the user problem, evidence it exists, why now
  2. Goals & Success Metrics — measurable targets with numbers and a measurement source (analytics event, funnel step). "Users are happier" is not a metric.
  3. Users & Scenarios — target segment(s) and 2-4 concrete usage scenarios
  4. Scope (MoSCoW) — Must / Should / Could / Won't for this release. The Won't list is mandatory — an empty Won't list means scope was not decided.
  5. Functional Requirements — numbered FR-1, FR-2, … testable statements
  6. Non-Functional Requirements — performance, security, accessibility (cite the committed tier), i18n, offline behavior where relevant
  7. Dependencies — other PRDs, ADRs, external services; each with a note on what breaks if the dependency slips
  8. Acceptance Criteria — checkable list a QA agent can verify; each criterion maps to at least one FR

After drafting all sections, show the user a compact summary of what will be written and ask for approval before writing the file:

"May I write this PRD to product/prd/prd-<feature>.md?"

Only write after explicit approval. Write the full document with frontmatter:

---
feature: <feature-name>
status: Draft   # Draft → In Review → Approved
version: 0.1
---

4. Verdict & Handoff

Self-check before closing (all must hold, else fix before reporting):

  • All 8 sections present, none placeholder-only
  • Every success metric has a number and a measurement source
  • Every acceptance criterion maps to an FR
  • Won't list is non-empty

Report the verdict:

  • COMPLETE — PRD written, all checks hold. Set status: In Review.
  • BLOCKED — a business fact is missing and the user must decide (name it).

Recommended next steps

  • /design-review product/prd/prd-<feature>.md — validate the PRD before it feeds architecture
  • /ux-design — author UX specs for this feature's key screens
  • More features to specify? Run /create-prd again — the catalog step is repeatable (one PRD per feature)
  • All MVP features covered? → /create-architecture (Architecture phase)

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.