Clarification Phase

1. Extract [NEEDS CLARIFICATION] markers from spec.md
2. Prioritize using matrix (Critical/High → ask, Medium/Low → assumptions)
3. Call AskUserQuestion with ≤3 questions (batched, multiSelect for subsystems)
4. Receive user answers synchronously
5. Integrate into spec.md, remove all markers
6. Add Clarifications (Resolved) section with deferred assumptions

Inputs: spec.md with [NEEDS CLARIFICATION] markers Outputs: Updated spec.md (no markers), clarifications.md (record)

• Spec phase completed (spec.md exists)
• spec.md contains ≥1 [NEEDS CLARIFICATION] marker (if none, skip /clarify)
• Git working tree clean

If clarification count >5, review spec phase quality (too many ambiguities).

Read spec.md, find all [NEEDS CLARIFICATION: ...] markers, extract ambiguity context.

# Count clarifications
grep -c "\[NEEDS CLARIFICATION" specs/NNN-slug/spec.md

# List with line numbers
grep -n "\[NEEDS CLARIFICATION" specs/NNN-slug/spec.md

If count = 0, skip /clarify phase.

Categorize each clarification by priority:

• Critical (always ask): Scope boundary, security/privacy, breaking changes
• High (ask if ambiguous): User experience decisions, functionality tradeoffs
• Medium (use defaults): Performance SLAs, technical stack choices
• Low (use standards): Error messages, rate limits

Keep only Critical + High priority questions (target: ≤3).

Convert Medium/Low to informed guesses, document as assumptions.

See references/prioritization-matrix.md for detailed categorization rules.

For each Critical/High priority clarification, structure as AskUserQuestion parameter:

AskUserQuestion format:

AskUserQuestion({
  questions: [
    {
      question:
        "spec.md:45 mentions 'dashboard metrics' but doesn't specify which. What should we display?",
      header: "Metrics", // max 12 chars
      multiSelect: false,
      options: [
        {
          label: "Completion only",
          description: "% of lessons finished (2 days, basic insights)",
        },
        {
          label: "Completion + time",
          description:
            "Lessons finished + hours logged (4 days, actionable insights)",
        },
        {
          label: "Full analytics",
          description:
            "Completion + time + quiz scores + engagement (7 days, requires infrastructure)",
        },
      ],
    },
  ],
});

Quality standards:

• question: Full context with spec.md reference (e.g., "spec.md:45 mentions...")
• header: Short label ≤12 chars (e.g., "Metrics", "Auth", "Scope")
• multiSelect: false for single choice, true for subsystems/features
• options: 2-3 concrete choices with impacts in description
• label: Concise option name (1-5 words)
• description: Implementation cost + value + tradeoffs (1-2 sentences)

Batch related questions (max 3 per AskUserQuestion call).

See references/question-bank.md for 40+ example questions in AskUserQuestion format.

For Medium/Low priority questions not asked, prepare assumptions section:

## Deferred Assumptions (Using Informed Guesses)

### [Topic]

**Not asked** (Low priority - standard default exists)
**Assumption**: [Concrete default choice]
**Rationale**: [Why this default is reasonable]
**Override**: [How user can override in spec.md]

These will be included in clarifications.md record after AskUserQuestion call.

Execute AskUserQuestion with batched Critical/High questions:

AskUserQuestion({
  questions: [
    {
      question:
        "spec.md:45 mentions 'dashboard metrics'. Which should we display?",
      header: "Metrics",
      multiSelect: false,
      options: [
        { label: "Completion only", description: "2 days, basic insights" },
        {
          label: "Completion + time",
          description: "4 days, actionable insights",
        },
        {
          label: "Full analytics",
          description: "7 days, requires infrastructure",
        },
      ],
    },
    {
      question:
        "spec.md:67 doesn't specify access control model. Which approach?",
      header: "Access",
      multiSelect: false,
      options: [
        {
          label: "Simple (users/admins)",
          description: "2 days, basic permissions",
        },
        {
          label: "Role-based (RBAC)",
          description: "4 days, flexible permissions",
        },
      ],
    },
  ],
});

Batching strategy:

• Batch 2-3 related questions per call
• Use multiSelect: true for subsystem/feature selection questions
• Use multiSelect: false for single-choice decisions

Tool returns answers object:

{
  "Metrics": "Completion + time",
  "Access": "Role-based (RBAC)"
}

User can also select "Other" for custom answers.

Use answers from AskUserQuestion tool response to update spec:

For each answered question:

1. Locate corresponding [NEEDS CLARIFICATION] marker in spec.md
2. Replace with concrete requirement based on selected option
3. Remove marker

Example:

// AskUserQuestion returned:
{
  "Metrics": "Completion + time",
  "Access": "Role-based (RBAC)"
}

Update spec.md:

<!-- Before -->

Dashboard displays student progress [NEEDS CLARIFICATION: Which metrics?]
Users can access dashboard [NEEDS CLARIFICATION: Access control?]

<!-- After -->

Dashboard displays:

- Lesson completion rate (% of assigned lessons finished)
- Time spent per lesson (hours logged)

User access control (role-based):

- Teachers: View assigned students only
- Admins: View all students
- Students: View own progress only

Validate integration:

# Must return 0 (no markers remain)
grep -c "\[NEEDS CLARIFICATION" specs/NNN-slug/spec.md

Generate specs/NNN-slug/clarifications.md as historical record:

# Clarifications for [Feature Name]

**Date**: [timestamp]
**Questions Asked**: 2 (Critical: 1, High: 1)
**Deferred**: 3 assumptions

## Questions & Answers

### Q1: Dashboard Metrics (Critical)

**Question**: spec.md:45 mentions 'dashboard metrics'. Which should we display?
**Options**: Completion only | Completion + time | Full analytics
**Selected**: Completion + time
**Rationale**: Balances actionable insights with implementation cost (4 days vs 7)

### Q2: Access Control (High)

**Question**: spec.md:67 doesn't specify access control model. Which approach?
**Options**: Simple (users/admins) | Role-based (RBAC)
**Selected**: Role-based (RBAC)
**Rationale**: Future-proof for additional roles

## Deferred Assumptions

### Export Format (Low)

**Not asked** - Standard default exists
**Assumption**: CSV format
**Rationale**: Most compatible, industry standard
**Override**: Specify in spec.md if JSON/Excel needed

### Rate Limiting (Low)

**Not asked** - Reasonable default
**Assumption**: 100 requests/minute per user
**Rationale**: Conservative, prevents abuse
**Override**: Specify in spec.md if higher limits needed

Add "Clarifications (Resolved)" section to spec.md:

## Clarifications (Resolved)

Answered 2 questions on [date]:

1. Dashboard metrics: Completion + time spent (4 days)
2. Access control: Role-based RBAC (future-proof)

Deferred assumptions: Export format (CSV), Rate limiting (100/min)

See clarifications.md for full details.

git add specs/NNN-slug/clarifications.md specs/NNN-slug/spec.md
git commit -m "docs: resolve clarifications for [feature-name]

Answered N questions:
- [Q1 summary]: [Decision]
- [Q2 summary]: [Decision]

Deferred assumptions:
- [Topic]: [Choice] ([reason])

All [NEEDS CLARIFICATION] markers removed
Ready for planning phase"

Update state.yaml: clarification.status = completed

• All [NEEDS CLARIFICATION] markers removed from spec.md
• ≤3 structured questions asked (Critical + High only)
• Medium/Low priorities documented as assumptions
• Answers integrated into spec.md Requirements section
• Clarifications (Resolved) section added to spec.md
• clarifications.md generated with user answers
• state.yaml updated (clarification.status = completed)
• Git commit created with descriptive message

Why: Delays workflow, frustrates users, causes analysis paralysis Target: ≤3 questions total after prioritization

Example (bad):

7 questions for export feature:
1. Export format? (CSV/JSON) → Has default ❌
2. Which fields? → Critical ✅
3. Email notification? → Has default ❌
4. Rate limiting? → Has default ❌
5. Max file size? → Has default ❌
6. Retention period? → Has default ❌
7. Compress files? → Has default ❌

Should be:

1 question (Critical):
- Which fields to export? (no reasonable default)

6 deferred assumptions:
- Format: CSV (standard)
- Email: Optional (user preference)
- Rate limit: 100/min (reasonable)
- Max size: 50MB (standard)
- Retention: 90 days (compliance standard)
- Compression: Auto >10MB (performance)

✅ Do: Use AskUserQuestion with clear context, 2-3 concrete options, quantified impacts

Why: Unclear questions lead to ambiguous answers, require follow-up, waste time

Example (good with AskUserQuestion):

AskUserQuestion({
  questions: [
    {
      question:
        "spec.md:45 mentions 'progress' but doesn't specify which metrics to display. What should the dashboard show?",
      header: "Metrics",
      multiSelect: false,
      options: [
        {
          label: "Completion only",
          description: "% of lessons finished (2 days, basic insights)",
        },
        {
          label: "Completion + time",
          description:
            "Lessons finished + hours logged (4 days, actionable insights for identifying struggling students)",
        },
        {
          label: "Full analytics",
          description:
            "Completion + time + quiz scores + engagement (7 days, requires analytics infrastructure)",
        },
      ],
    },
  ],
});

Result: Clear, specific options with quantified impacts - user can make informed decision.

Why: Planning phase can't proceed without concrete requirements in spec

Validation:

# Must return 0 (no markers remain)
grep -c "\[NEEDS CLARIFICATION" specs/NNN-slug/spec.md

Why: User doesn't know what defaults were applied, can't override if needed

Example:

## Deferred Assumptions

### Rate Limiting

**Not asked** (Low priority - reasonable default)
**Assumption**: 100 requests/minute per user
**Rationale**: Prevents abuse, can increase based on usage
**Override**: Specify in spec.md if higher limits needed

✅ Do: Provide 2-3 concrete options with quantified impacts

• Options: A/B/C with implementation costs + user value

Why: Open-ended answers are hard to integrate into spec, lead to follow-up questions

• question: Full context with spec.md reference (e.g., "spec.md:45 mentions...")
• header: Short label ≤12 chars (e.g., "Metrics", "Auth")
• options: 2-3 concrete choices with implementation costs in description
• multiSelect: false for single choice, true for subsystems/features

Example:

AskUserQuestion({
  questions: [
    {
      question: "spec.md:45 mentions 'metrics'. What should we display?",
      header: "Metrics",
      multiSelect: false,
      options: [
        { label: "Completion only", description: "2 days, basic" },
        { label: "Completion + time", description: "4 days, actionable" },
        { label: "Full analytics", description: "7 days, complex" },
      ],
    },
  ],
});

Result: Clear answers, faster decisions, easy spec integration

Target: ≤3 questions (Critical + High only)

Result: Focused user attention, faster responses, reasonable defaults

Result: Complete spec, ready for planning phase

• All [NEEDS CLARIFICATION] markers removed from spec.md (grep returns 0)
• ≤3 structured questions asked (Critical + High priority only)
• Medium/Low questions documented as assumptions with rationale
• User provided answers to all questions
• Answers integrated into spec.md Requirements section
• Clarifications (Resolved) section added to spec.md
• clarifications.md generated with questions + answers
• Deferred assumptions documented in clarifications.md
• Git commit created with descriptive message
• state.yaml updated (clarification.status = completed)
• spec.md is complete and unambiguous (ready for /plan)

• Question count: ≤3 per feature
• Question clarity: 100% follow structured format (Context → Options → Impact)
• Response integration: 100% (no remaining markers)
• Follow-up questions: <10%
• Time to resolution: ≤2 hours (excluding user response time)

Good clarifications:

• ≤3 questions (rigorously prioritized)
• Structured format (Context → Options → Impact → Recommendation)
• Concrete options (2-3 specific choices, not open-ended)
• Quantified impacts (implementation costs + user value)
• Clear recommendations (suggested option with rationale)
• Complete integration (all markers removed from spec)

Bad clarifications:

• 5 questions (didn't prioritize)

• Vague questions ("What should we do?")
• Compound questions (mixing multiple decisions)
• No options (open-ended)
• Missing integration (markers remain in spec)

Issue: Questions are vague Solution: Use AskUserQuestion format with clear context + spec reference, 2-3 concrete options, quantified impacts in description

Issue: User can't choose between options Solution: Add more context to question text, include cost/benefit tradeoffs in option descriptions

Issue: AskUserQuestion header too long Solution: Keep header ≤12 chars (e.g., "Metrics" not "Dashboard Metrics Scope")

Issue: [NEEDS CLARIFICATION] markers remain after integration Solution: Extract answers from AskUserQuestion response, update spec.md for each marker, run validation check

Issue: Planning phase blocked due to ambiguity Solution: Spec integration incomplete, verify answers mapped to spec requirements correctly

See templates/ for:

• clarification-template.md (clarifications.md record template)