requirements-interview
Requirements Interview
Run structured, multi-round requirement interviews, then convert answers into actionable documents.
When to Use
Use this skill when:
- Requirements are incomplete, ambiguous, or contradictory.
- The user asks to "clarify requirements" through Q&A.
- You need interview-driven outputs (SPEC/PRD/PLAN), not only a static PRD.
Do not use this skill when:
- The user already provided clear and complete requirements.
- The user only needs a standard PRD summary (use
product-manager).
Inputs
- Existing
SPEC.md(optional) - Existing
PRD.md(optional) - Existing
PLAN.md(optional)
Tool Strategy
Primary: AskUserQuestionTool (or equivalent interactive question tool).
Fallback: If the tool is unavailable, continue with plain conversational questions and keep the same interview structure.
Core Rules
- Ask deep, non-trivial questions (avoid obvious yes/no prompts).
- Minimum 2 rounds of clarification unless user explicitly stops.
- After each round, update assumptions and unresolved items.
- Separate confirmed facts from hypotheses.
- Do not skip final document outputs.
Interview Dimensions (Mandatory)
Cover these areas:
- Business goals and success metrics
- Target users and key scenarios
- Scope boundaries and explicit non-goals
- UX expectations and acceptance criteria
- Technical constraints and integration boundaries
- Security/compliance/performance requirements
- Risks, tradeoffs, and alternatives
- Milestones, priority, and delivery sequence
Workflow
- Read existing
SPEC.md/PRD.md/PLAN.mdif they exist. - Summarize known facts and gaps.
- Run interview round 1 with focused open questions.
- Update SPEC/PRD/PLAN drafts.
- Run interview round 2+ until requirements are actionable.
- Produce final deliverables and a concise uncertainty log.
Outputs
Always provide or update:
SPEC.md: problem definition, scope, constraints, acceptance checksPRD.md: user value, requirements, priorities, success metricsPLAN.md: implementation phases, dependencies, risks, test strategy
Completion Criteria
- Requirements are implementation-ready.
- Key assumptions and tradeoffs are explicit.
- Acceptance criteria are testable.
- Open questions are listed with owner and next action.
Handoff Guidance
If implementation is next:
- handoff to
dev-plannerfor execution planning, or - handoff to
product-managerfor roadmap-level prioritization.
Examples
Example 1: Ambiguous Feature Request
User says: "I want to add user authentication" Actions:
- Ask about auth method (OAuth, password-based, SSO?)
- Clarify user roles and permissions
- Determine session management preferences
- Define password reset and recovery flows Result: Complete auth SPEC.md with acceptance criteria
Example 2: New Project Startup
User says: "Help me define requirements for a CRM system" Actions:
- Identify target users (sales, support, management?)
- Define core workflows (contact management, deals, tasks)
- Clarify integrations (email, calendar, phone)
- Establish success metrics Result: Full PRD.md with prioritized features
Example 3: Scope Refinement
User says: "The current feature is too complex, simplify it" Actions:
- List current scope items
- Identify which are must-have vs nice-to-have
- Propose simplified scope with clear non-goals
- Update SPEC.md boundaries Result: Streamlined specification with explicit exclusions
More from jochenyang/jochen-ai-rules
ui-ux-pro-max
UI/UX design-system reasoning and UX quality audit skill. Use when user needs style direction, palette/typography selection, UX review, or design optimization before implementation. Do NOT use for backend logic or database design.
24devops-engineer
CI/CD pipeline design, containerization, and infrastructure management. Handles Docker, Kubernetes, monitoring setup (Prometheus/Grafana), and infrastructure-as-code (Terraform/Pulumi). Use when user asks to deploy, configure CI/CD, set up Docker/K8s, or manage infrastructure.
19handoff
Create and resume structured manual session handoffs for long-running development work. Use when approaching context limits, before manual reset, before switching models or IDEs, after a milestone, or when automatic compact would lose important implementation state.
1reflect
Review current conversation, analyze tasks, errors, and user feedback, extract learning opportunities for skill improvement. Use when user says "reflect", "review session", "what did we learn", "session summary", or after completing a complex task.
1developer
Comprehensive full-stack development for web, mobile, and game projects. Handles frontend (React/Vue/Angular), backend (Node.js/Python/Go/Java), mobile (Flutter/React Native/Swift/Kotlin), and game development (Unity/Unreal/Godot). Use when user asks to build, create, develop, implement, debug, or fix any web, mobile, or game project. Do NOT use for design-only tasks (use frontend-design or ui-ux-pro-max instead).
1mcp-builder
MCP server development for AI agents. Designs tool schemas, implements Python/TypeScript servers, creates evaluation tests. Use when user asks to build MCP server, create tool integration, or develop Claude plugins. Supports GitHub/Notion/Slack integrations.
1