Token导航 LogoToken导航TokenDH.com
研究检索需要联网github未标认证来源可访问许可证需确认审计未展示

ring%3apre-dev-prd-creationring%3apre dev prd 创建

Agent Skill

ring%3apre-dev-prd-creation 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

870

周安装

37

GitHub Stars

180

下载量

305
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:ring%3apre-dev-prd-creation(ring%3apre dev prd 创建)
来源仓库:https://github.com/lerianstudio/ring
仓库路径:skills/ring%3Apre-dev-prd-creation
安装命令:
npx skills add https://github.com/lerianstudio/ring --skill ring:pre-dev-prd-creation
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/lerianstudio/ring --skill ring:pre-dev-prd-creation

简介

ring%3apre-dev-prd-creation 用于查找、检索和筛选相关信息。

  • 它适合在 Codex、Claude、Cursor、Gemini CLI 中基于关键词或场景快速获取资料。
  • 可通过 npx skills add 命令从 GitHub 仓库安装,详细用法见原始文档。
  • 使用前应核实权限、维护情况及是否涉及网络或文件操作。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

PRD Creation - Business Before Technical

Foundational Principle

Business requirements (WHAT/WHY) must be fully defined before technical decisions (HOW/WHERE).

Mixing business and technical concerns creates:

  • Requirements that serve implementation convenience, not user needs
  • Technical constraints that limit product vision
  • Inability to evaluate alternatives objectively
  • Cascade failures when requirements change

The PRD answers: WHAT we're building and WHY it matters to users and business. The PRD never answers: HOW we'll build it or WHERE components will live.

Mandatory Workflow

PhaseActivities
0. Load ResearchCheck docs/pre-dev/{feature}/research.md; review codebase patterns, best practices, framework constraints, UX research; reference findings with file:line notation
1. Problem DiscoveryDefine problem without solution bias; identify specific users; quantify pain with metrics/evidence
2. Business RequirementsExecutive summary (3 sentences); user personas (goals, frustrations); user stories (As/I want/So that); success metrics (measurable); scope boundaries (in/out)
3. Gate 1 ValidationProblem articulated; impact quantified; users identified; features address problem; metrics measurable; scope explicit
4. UX ValidationDispatch product-designer to validate PRD against user needs and create ux-criteria.md

Explicit Rules

✅ DO Include in PRD

Problem definition and user pain points, user personas (demographics, goals, frustrations), user stories with acceptance criteria, feature requirements (WHAT not HOW), success metrics (adoption, satisfaction, KPIs), scope boundaries (in/out explicitly), go-to-market considerations

❌ NEVER Include in PRD

Architecture diagrams or component design, technology choices (languages, frameworks, databases), implementation approaches or algorithms, database schemas or API specifications, code examples or package dependencies, infrastructure needs or deployment strategies, system integration patterns

Separation Rules

  1. If it's a technology name → Not in PRD (goes in Dependency Map)
  2. If it's a "how to build" → Not in PRD (goes in TRD)
  3. If it's implementation → Not in PRD (goes in Tasks/Subtasks)
  4. If it describes system behavior → Not in PRD (goes in TRD)

Rationalization Table

ExcuseReality
"Just a quick technical note won't hurt"Technical details constrain business thinking. Keep them separate.
"Stakeholders need to know it's feasible"Feasibility comes in TRD after business requirements are locked.
"The implementation is obvious"Obvious to you ≠ obvious to everyone. Separate concerns.
"I'll save time by combining PRD and TRD"You'll waste time rewriting when requirements change.
"This is a simple feature, no need for formality"Simple features still need clear requirements. Follow the process.
"I can skip Gate 1, I know it's good"Gates exist because humans are overconfident. Validate.
"The problem is obvious, no need for personas"Obvious to you ≠ validated with users. Document it.
"Success metrics can be defined later"Defining metrics later means building without targets. Do it now.
"I'll just add this one API endpoint detail"API design is technical architecture. Stop. Keep it in TRD.
"But we already decided on PostgreSQL"Technology decisions come after business requirements. Wait.
"CEO/CTO says it's a business constraint"Authority doesn't change what's technical. Abstract it anyway.
"Investors need to see specific vendors/tech"Show phasing and constraints abstractly. Vendors go in TRD.
"This is product scoping, not technical design"Scope = capabilities. Technology = implementation. Different things.
"Mentioning Stripe shows we're being practical"Mentioning "payment processor" shows the same. Stay abstract.
"PRDs can mention tech when it's a constraint"PRDs mention capabilities needed. TRD maps capabilities to tech.
"Context matters - this is for exec review"Context doesn't override principles. Executives get abstracted version.

Security Requirements Discovery (Business Level)

During PRD creation, identify if the feature requires access control:

Business QuestionIf Yes → Document
Does this feature handle user-specific data?"Users can only access their own [data type]"
Are there different user roles with different permissions?"Admins can [X], regular users can [Y]"
Does this feature need to identify who performed an action?"Audit trail required for [action type]"
Does this integrate with other internal services?"Service must authenticate to [service name]"
Are there regulatory requirements (GDPR, PCI-DSS, HIPAA)?"Must comply with [regulation] for [data type]"

What to include in PRD:

  • ✅ "Only authenticated users can access this feature"
  • ✅ "Users can only view/edit their own records"
  • ✅ "Admin approval required for [action]"
  • ✅ "Must track who performed each action"

What NOT to include in PRD:

  • ❌ "Use JWT tokens" (technology choice → TRD)
  • ❌ "Integrate with Access Manager" (architecture → TRD)
  • ❌ "OAuth2 flow" (protocol choice → TRD)

Note: The TRD (Gate 3) will translate these business requirements into authentication/authorization architecture patterns. For Go services, refer to golang.md → Access Manager Integration section during TRD creation.


Operational Dashboard Discovery (Business Level)

During PRD creation, assess whether the feature/product requires business dashboard endpoints.

Not every product needs dashboard endpoints. The deciding factor is whether a human operator or business manager will need aggregated visibility to make decisions.

When to Ask

This discovery MUST be triggered when:

  • Defining a new product or major feature
  • The feature involves data that accumulates over time (transactions, events, operations)
  • The feature is operated by end-users (not purely infrastructure)

Discovery Question

Use AskUserQuestion tool:

Question: "Will an operator or business manager need a consolidated view of this feature's data to make decisions (e.g., volumes, statuses, error rates, trends)?"

  • Header: "Operational Dashboard"
  • multiSelect: false
  • Options:

- "Yes - Business dashboard needed" (description: "Operators need aggregated metrics, trends, and status overviews to manage this feature day-to-day") - "No - Infrastructure/backend only" (description: "This feature runs silently — monitoring is handled by observability tooling, not business dashboards") - "Not sure - Needs discussion" (description: "Flag for product/business review before deciding")

If "Yes" Selected

Document in PRD under a "Dashboard Requirements" section:

QuestionDocument
Who is the primary dashboard consumer?Persona: operator, manager, compliance officer, etc.
What decisions does the dashboard support?"Is this healthy?", "Where are bottlenecks?", "Are we meeting SLAs?"
What key metrics need visibility?Volumes, rates, distributions, trends, aging
What is the expected refresh cadence?Real-time, near real-time, periodic

Example PRD section:

## Dashboard Requirements

**Consumer:** Operations manager
**Decisions supported:** Health monitoring, SLA compliance, exception triage
**Key metrics:**
- Total volume processed (by period)
- Distribution by status (success/pending/failed)
- Error rate trends
- Aging of unresolved items
**Refresh:** Near real-time (< 5 min lag)

This section cascades into Feature Map (Gate 2) for UI components and API Design (Gate 4) for dashboard data endpoints.

If "No" Selected

Document explicitly in PRD scope:

## Out of Scope
- Business dashboard endpoints (infrastructure product — monitoring via observability stack)

This prevents the question from resurfacing in later gates.

If "Not Sure" Selected

Flag as open question in PRD:

## Open Questions
- [ ] Dashboard requirements TBD — needs product/business review before Gate 2

STOP Gate 1 validation for this item until resolved.

Guidance: When Dashboards Make Sense

Product TypeDashboard?Rationale
Core ledger / transaction engineYesOperators need volume, status, health visibility
Reconciliation / matchingYesException management requires aggregated view
Risk / fraud analysisYesMonitoring product by nature — alerts, scores, trends
Regulatory reportingYesCompliance needs deadline tracking, coverage, delivery status
Payment rails (PIX, SWIFT, etc.)YesHigh volume + SLA requirements = operational visibility
Billing / feesYesRevenue data needs consolidated view
Workflow orchestrationMaybeDepends on volume and operational complexity
Data ingestionNoBackend plumbing — health is infra/o11y concern
Auth / identityNoInfrastructure — login metrics are o11y, not business dashboard
Tenant managementNoAdmin/platform tooling, not end-user operated

Rationalization Table

ExcuseReality
"We'll add dashboards later"Later = never, or expensive rework. Decide now.
"Every product needs a dashboard"Infrastructure products don't. Forcing it creates noise.
"Observability covers it"O11y ≠ business dashboard. Grafana for SREs ≠ console for operators.
"Dashboard is a frontend concern"Dashboard data endpoints are backend. Decide scope here.
"It's obvious this needs a dashboard"Obvious to you ≠ documented. Ask and record the decision.

Red Flags - STOP

If you catch yourself writing or thinking any of these in a PRD, STOP:

  • Technology product names (PostgreSQL, Redis, Kafka, AWS, etc.)
  • Framework or library names (React, Fiber, Express, etc.)
  • Words like: "architecture", "component", "service", "endpoint", "schema"
  • Phrases like: "we'll use X to do Y" or "the system will store data in Z"
  • Code examples or API specifications
  • "How we'll implement" or "Technical approach"
  • Database table designs or data models
  • Integration patterns or protocols

When you catch yourself: Move that content to a "technical notes" section to transfer to TRD later. Keep PRD pure business.

Data Source Discovery (Frontend-only Features)

⛔ MANDATORY: If feature is frontend-only (uses existing backend APIs), this section MUST be completed.

When to Apply

Check research.md frontmatter for topology:

topology:
  scope: frontend-only  # ← This triggers data source discovery

Step 1: Identify Existing APIs

Document all existing APIs the feature will consume:

## Data Sources

### Existing Backend APIs

| API | Endpoint Pattern | Description | Documentation |
|-----|------------------|-------------|---------------|
| User API | /api/v1/users/* | User management | link to docs |
| Orders API | /api/v1/orders/* | Order operations | link to docs |

### API Capabilities Needed

| User Story | Required Capability | Available API | Gap? |
|------------|---------------------|---------------|------|
| US-001 | Get user profile | GET /users/:id | No |
| US-002 | List user orders | GET /orders?userId= | No |
| US-003 | Export order PDF | None | Yes - needs BFF |

Step 2: Identify API Gaps

If user story requires capability not available in existing APIs:

Gap TypeAction
Data aggregation neededFlag: "BFF required for aggregation"
Data transformation neededFlag: "BFF required for transformation"
Missing endpointFlag: "Backend enhancement needed"
Multiple API calls for single viewFlag: "BFF recommended for optimization"

Step 3: Document in PRD

Add to PRD under "Technical Context" section:

## Technical Context (Frontend-only)

**Data Source Type:** Existing Backend APIs

**Available APIs:**
- User API (v1) - user management operations
- Orders API (v1) - order CRUD operations

**API Gaps Identified:**
- [ ] No endpoint for aggregated dashboard data → BFF needed
- [ ] PDF export not available → Backend enhancement OR BFF generation

**BFF Requirements:** [None | Aggregation | Transformation | Both]

Rationalization Table for Data Source Discovery

ExcuseReality
"We'll discover APIs during implementation"Discovery during implementation causes rework. Document now.
"Frontend devs know the APIs"Documentation prevents tribal knowledge. Write it down.
"APIs are obvious from the codebase"Obvious to you ≠ documented for AI agents. Be explicit.
"We don't need BFF, just call APIs directly"Multiple API calls = poor UX. Evaluate BFF need properly.
"BFF adds complexity"BFF complexity < spaghetti frontend API calls. Evaluate objectively.

Gate 1 Validation Checklist

CategoryRequirements
Problem DefinitionProblem articulated (1-2 sentences); impact quantified/qualified; users specifically identified; current workarounds documented
Solution ValueFeatures address core problem; success metrics measurable; ROI case documented; user value clear per feature
Scope ClarityIn-scope items explicit; out-of-scope with rationale; assumptions documented; business dependencies identified
Market FitDifferentiation clear; value proposition validated; business case sound; go-to-market outlined
Data Sources (frontend-only)Existing APIs documented; API capabilities mapped to user stories; API gaps identified; BFF requirements determined

Gate Result: ✅ PASS → UX Validation → Feature Map | ⚠️ CONDITIONAL (address gaps) | ❌ FAIL (return to discovery)

Phase 4: UX Validation + Wireframes

After PRD passes Gate 1 validation, dispatch product-designer for UX validation:

Task(
  subagent_type="ring:product-designer",
  prompt="Validate PRD at docs/pre-dev/{feature}/prd.md against user needs. Mode: ux-validation.

  UI Configuration (from pre-dev command):
  - UI Library: {ui_library}  // e.g., shadcn/ui, Chakra UI, or auto-detected
  - Styling: {styling}        // e.g., TailwindCSS, CSS Modules, or auto-detected

  Create ux-criteria.md with: problem validation status, refined personas, UX acceptance criteria (functional, usability, accessibility, responsive).

  If feature has UI components, also create wireframes/ directory with low-fidelity prototypes using the specified UI library components and styling approach."
)

IMPORTANT: Pass UI Configuration to product-designer

The UI Library and Styling choices (from /pre-dev-feature or /pre-dev-full questions) MUST be passed to product-designer:

  • If user selected a library → Use that library's component names in wireframes
  • If auto-detected from package.json → Use the detected library
  • If "Custom components only" → Use generic component names

This ensures wireframes reference real components that will be available during implementation.

UX Validation Outputs:

  • docs/pre-dev/{feature}/ux-criteria.md - UX acceptance criteria
  • docs/pre-dev/{feature}/wireframes/ - Low-fidelity prototypes (if feature has UI)

- {screen-name}.yaml - YAML wireframe specification per screen - user-flows.md - User flow diagrams with state transitions

UI Detection Rule: If PRD contains any of these, feature HAS UI and wireframes are REQUIRED:

  • User stories mentioning: "see", "view", "click", "navigate", "page", "screen", "button", "form"
  • Features involving: login, dashboard, settings, profile, reports, notifications
  • Any user-facing interaction

UX Validation Checklist:

CheckRequiredCondition
Problem validation status documentedYesAlways
Personas refined based on PRDYesAlways
Functional UX criteria definedYesAlways
Usability criteria definedYesAlways
Accessibility criteria definedYesAlways
Responsive criteria definedYesAlways
All PRD user stories have UX criteriaYesAlways
Wireframes created for each screenYesIf feature has UI
User flows documentedYesIf feature has UI
State coverage table completeYesIf feature has UI

Wireframe YAML Format:

screen: Screen Name
route: /path
layout: layout-type
components:
  - id: component-id
    type: component-type
    # ... component specs
states:
  default: { ... }
  loading: { ... }
  error: { ... }
responsive:
  mobile: { ... }
  desktop: { ... }
accessibility:
  keyboard: [ ... ]
  screen-reader: [ ... ]

Document Placement (based on topology.structure)

prd.md placement:

Structureprd.md Location
single-repodocs/pre-dev/{feature}/prd.md
monorepodocs/pre-dev/{feature}/prd.md (root)
multi-repoWrite to BOTH repos

ux-criteria.md and wireframes/ placement:

StructureLocation
single-repodocs/pre-dev/{feature}/
monorepo{frontend.path}/docs/pre-dev/{feature}/
multi-repo{frontend.path}/docs/pre-dev/{feature}/

Why frontend path for UX docs? UX criteria and wireframes are consumed by frontend engineers. Placing them in the frontend module/repo ensures they are discoverable where they'll be used.

Directory creation for multi-module:

# Read topology from research.md frontmatter
# Create appropriate directories:

# For monorepo - frontend module
mkdir -p "{frontend.path}/docs/pre-dev/{feature}"

# For multi-repo - both repos for prd.md, frontend for UX
mkdir -p "{backend.path}/docs/pre-dev/{feature}"
mkdir -p "{frontend.path}/docs/pre-dev/{feature}"

If UX validation fails:

  • Conflicting user needs → Return to Phase 1 (Problem Discovery)
  • Missing persona details → Enrich PRD personas
  • Unclear acceptance criteria → Iterate with product-designer
  • Missing wireframes for UI feature → product-designer must create them

Common Violations

ViolationWrongCorrect
Tech in Features"FR-001: Use JWT tokens for session, bcrypt for passwords, OAuth2 with Google""FR-001: Users can create accounts and securely log in. Value: Access personalized content. Success: 95% authenticate first attempt"
Implementation in Stories"As user, I want to store data in PostgreSQL so queries are fast""As user, I want dashboard to load in <2 seconds so I can quickly access information"
Architecture in Problem"Our microservices architecture doesn't support real-time notifications""Users miss important updates because they must manually refresh. 78% report missing time-sensitive info"
Authority-Based Bypass"MVP: Stripe for payments, PostgreSQL (we already use it)""Phase 1: Integrate with existing payment vendor (2-week timeline); leverage existing database infrastructure. TRD will document specific vendor selection"

Confidence Scoring

FactorPointsCriteria
Market Validation0-25Direct user feedback: 25, Market research: 15, Assumptions: 5
Problem Clarity0-25Quantified pain: 25, Qualitative evidence: 15, Hypothetical: 5
Solution Fit0-25Proven pattern: 25, Adjacent pattern: 15, Novel: 5
Business Value0-25Clear ROI: 25, Indirect value: 15, Uncertain: 5

Action: 80+ autonomous | 50-79 present options | <50 ask discovery questions

Design System Generation (For New Projects with UI)

MANDATORY: If feature has UI (Q4=Yes) AND project is new (no existing design system), generate design-system.md based on Q7-Q10 responses.

Trigger Conditions:

  • Q4 = "Yes" (feature has UI)
  • No existing globals.css with CSS variables OR no tailwind.config.* with custom colors
  • Q7-Q10 were answered (not auto-detected from existing config)

design-system.md Template:

# Design System - {Feature Name}

## Configuration Source
- Accessibility: {Q7 response}
- Dark Mode: {Q8 response}
- Brand Color: {Q9 response}
- Typography: {Q10 response}

## Color Palette

### Primary
| Token | Light Mode | Dark Mode | Usage |
|-------|------------|-----------|-------|
| `--primary` | {derived from Q9} | {lighter for dark} | Main actions, links |
| `--primary-foreground` | #ffffff | {dark text} | Text on primary |
| `--primary-hover` | {darker shade} | {lighter shade} | Hover states |

### Neutral
| Token | Light Mode | Dark Mode | Usage |
|-------|------------|-----------|-------|
| `--background` | #ffffff | #0f172a | Page background |
| `--foreground` | #0f172a | #f8fafc | Primary text |
| `--muted` | #f1f5f9 | #1e293b | Secondary backgrounds |
| `--muted-foreground` | #64748b | #94a3b8 | Secondary text |
| `--border` | #e2e8f0 | #334155 | Borders, dividers |

### Semantic
| Token | Light Mode | Dark Mode | Usage |
|-------|------------|-----------|-------|
| `--success` | #16a34a | #22c55e | Success states |
| `--warning` | #ca8a04 | #eab308 | Warning states |
| `--error` | #dc2626 | #ef4444 | Error states |

## Contrast Validation ({Q7 level})

| Combination | Required Ratio | Actual | Pass |
|-------------|----------------|--------|------|
| foreground on background | {4.5:1 or 7:1} | {calculated} | ✅/❌ |
| primary on background | {4.5:1 or 7:1} | {calculated} | ✅/❌ |
| muted-foreground on background | {4.5:1 or 7:1} | {calculated} | ✅/❌ |

## Typography

### Font Stack
- Display: {Q10 choice}, sans-serif
- Body: {Q10 choice}, sans-serif
- Mono: 'Geist Mono', ui-monospace, monospace

### Scale
| Token | Size | Line Height | Usage |
|-------|------|-------------|-------|
| text-xs | 12px | 16px | Captions |
| text-sm | 14px | 20px | Secondary |
| text-base | 16px | 24px | Body |
| text-lg | 18px | 28px | Large body |
| text-xl | 20px | 28px | Small headings |
| text-2xl | 24px | 32px | Section headings |
| text-3xl | 30px | 36px | Page headings |

## Spacing Scale (4px base)

| Token | Value |
|-------|-------|
| spacing-1 | 4px |
| spacing-2 | 8px |
| spacing-3 | 12px |
| spacing-4 | 16px |
| spacing-6 | 24px |
| spacing-8 | 32px |

## Accessibility Requirements

- **Level:** {Q7 response}
- **Minimum contrast:** {4.5:1 for AA, 7:1 for AAA}
- **Focus indicators:** 2px solid ring required
- **Touch targets:** Minimum 44x44px
- **Reduced motion:** Respect prefers-reduced-motion

Color Derivation from Q9 (Brand Color):

Q9 Choice--primary (Light)--primary (Dark)
Bluehsl(217, 91%, 60%)hsl(217, 91%, 65%)
Purplehsl(262, 83%, 58%)hsl(262, 83%, 63%)
Greenhsl(142, 76%, 36%)hsl(142, 71%, 45%)
Orangehsl(25, 95%, 53%)hsl(25, 95%, 58%)
CustomUser-provided hexLightened 5%

Typography Mapping from Q10:

Q10 ChoiceFont Family
Modern Tech (Geist)'Geist', sans-serif
Contemporary (Satoshi)'Satoshi', sans-serif
Editorial (Cabinet Grotesk)'Cabinet Grotesk', sans-serif
Professional (General Sans)'General Sans', sans-serif

design-system.md Placement:

StructureLocation
single-repodocs/pre-dev/{feature}/design-system.md
monorepo{frontend.path}/docs/pre-dev/{feature}/design-system.md
multi-repo{frontend.path}/docs/pre-dev/{feature}/design-system.md

GATE BLOCKER: If feature has UI and project is new, design-system.md MUST exist before proceeding to TRD. The TRD will reference these tokens.

Output & After Approval

Outputs (paths depend on topology.structure):

Documentsingle-repomonorepomulti-repo
prd.mddocs/pre-dev/{feature}/docs/pre-dev/{feature}/Both repos
ux-criteria.mddocs/pre-dev/{feature}/{frontend.path}/docs/pre-dev/{feature}/Frontend repo
wireframes/docs/pre-dev/{feature}/{frontend.path}/docs/pre-dev/{feature}/Frontend repo
design-system.mddocs/pre-dev/{feature}/{frontend.path}/docs/pre-dev/{feature}/Frontend repo (new projects)
  1. ✅ Lock the PRD - no changes without formal amendment
  2. ✅ Lock ux-criteria.md - defines UX acceptance for implementation
  3. ✅ Lock wireframes/ - defines visual structure for ui-engineer
  4. 🎯 Use all as input for Feature Map (ring:pre-dev-feature-map) or TRD (ring:pre-dev-trd-creation)
  5. 🚫 Never add technical details retroactively
  6. 📋 Keep business/technical strictly separated

The Bottom Line

If you wrote a PRD with technical details, delete it and start over.

The PRD is business-only. Period. No exceptions. No "just this once". No "but it's relevant".

Technical details go in TRD. That's the next phase. Wait for it.

Follow the separation. Your future self will thank you.


Standards Loading (MANDATORY)

This skill is a business requirements skill and does NOT require WebFetch of language-specific standards.

Purpose: PRD defines WHAT/WHY at a business level. Technical standards are irrelevant at this stage—they apply during TRD (Gate 3) and implementation.

However, if research.md exists from Gate 0, MUST load and reference it for:

  • Existing codebase patterns
  • Prior solutions from docs/solutions/
  • Framework constraints discovered during research

Blocker Criteria - STOP and Report

ConditionActionSeverity
Problem statement cannot be articulatedSTOP and conduct discovery with stakeholdersCRITICAL
No user personas can be identifiedSTOP and validate who the feature servesCRITICAL
Technical details creep into PRDSTOP and remove them—move to TRD notesHIGH
Success metrics cannot be definedSTOP and clarify business valueHIGH
Scope boundaries unclear (in/out not defined)STOP and establish explicit scopeMEDIUM
research.md exists but wasn't consultedContinue but MUST reference research findingsMEDIUM

Cannot Be Overridden

These requirements are NON-NEGOTIABLE:

  • MUST NOT include technical implementation details in PRD
  • MUST NOT include technology names, frameworks, or databases
  • MUST define measurable success metrics
  • MUST include explicit scope boundaries (in-scope/out-of-scope)
  • MUST include user personas with goals and frustrations
  • MUST dispatch ring:product-designer for UX validation after PRD
  • CANNOT proceed to Gate 2/3 with unvalidated PRD

Severity Calibration

SeverityDefinitionExample
CRITICALPRD cannot be writtenNo problem identified, no users defined
HIGHPRD violates separation principleTechnology names in feature requirements
MEDIUMPRD incomplete but usableMissing 1-2 success metrics
LOWMinor quality issuesUser story format inconsistent

Pressure Resistance

User SaysYour Response
"Include the tech stack, stakeholders need it""Cannot include technology in PRD. Stakeholders see tech choices in TRD/Dependency Map. PRD stays business-only."
"Just mention PostgreSQL as a constraint""Cannot mention specific databases. Use 'persistent data storage' if needed. Tech selection happens in Gate 6."
"Skip UX validation, designer approved verbally""Cannot skip UX validation. Verbal approval isn't documented. I'll dispatch product-designer for formal validation."
"Success metrics can be defined later""Cannot defer metrics. Undefined metrics = building without targets. I'll define measurable KPIs now."
"Scope is obvious, skip boundaries""Cannot skip scope boundaries. 'Obvious' scope causes creep. I'll document explicit in/out lists."

Anti-Rationalization

RationalizationWhy It's WRONGRequired Action
"Quick technical note won't hurt"Technical details constrain business thinkingRemove all tech, move to TRD notes
"Stakeholders need feasibility context"Feasibility comes in TRD after requirements lockedKeep PRD pure business
"The implementation is obvious"Obvious to you ≠ obvious to everyoneSeparate concerns regardless
"Combining PRD and TRD saves time"Combining wastes time when requirements changeKeep documents separate
"This is a simple feature, no formality needed"Simple features still need clear requirementsFollow the process
"Problem is obvious, skip personas"Obvious to you ≠ validated with usersDocument personas
"Adding one API detail helps clarity"API design is TRD territory. Stop. Keep it in TRDRemove, add to TRD notes

When This Skill Is Not Needed

  • PRD already exists and is validated (Gate 1 passed)
  • Pure technical task without business impact (TRD directly)
  • Bug fix with known solution (use systematic-debugging)
  • Documentation-only updates
  • Configuration changes with no feature impact
  • Refactoring with no user-facing changes

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

36.7%
按下载量换算112

Claude

29.37%
按下载量换算90

Cursor

19.85%
按下载量换算61

Gemini CLI

10.18%
按下载量换算31

安全审计

暂无安全审计结果可展示。

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills