interview

Interview Command

Conduct a structured requirements elicitation interview using AI-led LLMREI patterns.

Usage

/requirements-elicitation:interview "Product Manager" --context "checkout redesign"
/requirements-elicitation:interview "End User" --mode guided --domain "e-commerce"
/requirements-elicitation:interview "Technical Lead" --mode semi-auto

Arguments

Argument	Required	Description
stakeholder-role	Yes	Role of the person being interviewed (e.g., "Product Manager", "End User")
--context	No	Project or feature context for the interview
--mode	No	Autonomy mode: guided, semi-auto, full-auto (default: semi-auto)
--domain	No	Domain name for organizing output files

Workflow

Step 1: Initialize Interview Context

Parse arguments and set up interview:

interview_setup:
  stakeholder_role: "{from argument}"
  project_context: "{from --context or ask}"
  autonomy_mode: "{from --mode or default}"
  domain: "{from --domain or derive from context}"
  session_id: "INT-{timestamp}"

Step 2: Load Interview Skill

Invoke the requirements-elicitation:interview-conducting skill to load question pathways and structure.

Step 3: Determine Interview Mode

If Real Stakeholder (default):

• Stakeholder is the human in the chat
• Use AskUserQuestion for structured questions
• Allow natural conversation flow

If Simulated (--simulate flag or no human response):

• Spawn appropriate persona agent
• Conduct interview with simulated stakeholder
• Mark all requirements with lower confidence

Step 4: Conduct Interview

Spawn the requirements-interviewer agent with the interview context.

The agent will:

1. Open the interview and establish rapport
2. Gather context about the stakeholder's role
3. Explore requirements systematically
4. Validate understanding through summarization
5. Close with next steps

Step 5: Save Results

Save interview results to:

.requirements/{domain}/interviews/INT-{session-id}.yaml

Step 6: Report Summary

Display:

• Number of requirements elicited
• Key themes identified
• Gaps or follow-up items
• Path to saved interview file

Examples

Basic Interview

/requirements-elicitation:interview "Product Manager"

Starts a semi-autonomous interview with a Product Manager, asking for project context.

Guided Interview with Context

/requirements-elicitation:interview "Security Officer" --context "authentication system" --mode guided

Starts a guided interview where each question is approved before asking.

Full-Auto Interview for Exploration

/requirements-elicitation:interview "Developer" --mode full-auto --domain "api-redesign"

Conducts a complete interview autonomously, presenting the summary at the end.

Output

During Interview

For guided mode:

AI: "I suggest asking: 'What security concerns do you have about the current system?'
     Do you approve this question?"

For semi-auto mode:

AI: [Conducts segment]
    "Interview progress: Context gathering complete.
     3 requirements identified so far.
     Ready to explore functional requirements. Continue?"

After Interview

Interview Complete: INT-20251225-143022

Stakeholder: Product Manager
Duration: ~25 minutes
Mode: semi-auto

Requirements Elicited: 12
  - Functional: 8
  - Non-Functional: 3
  - Constraints: 1

Key Themes:
  - User authentication improvements
  - Performance during peak hours
  - Mobile responsiveness

Gaps Identified: 2
  - Accessibility requirements not discussed
  - Disaster recovery needs clarification

Saved to: .requirements/checkout/interviews/INT-20251225-143022.yaml

Next Steps:
  - Review extracted requirements
  - Schedule follow-up for gaps
  - Run /requirements-elicitation:gaps for completeness check

Integration

Follow-Up Commands

After an interview:

# Check for gaps in elicited requirements
/requirements-elicitation:gaps

# Simulate other stakeholder perspectives
/requirements-elicitation:simulate "security" --personas technical,compliance

# Research domain-specific requirements
/requirements-elicitation:research "PCI-DSS compliance"

# Export to specification format
/requirements-elicitation:export --to canonical

With Multiple Stakeholders

Conduct multiple interviews, then consolidate:

/requirements-elicitation:interview "Product Manager" --domain "checkout"
/requirements-elicitation:interview "Developer" --domain "checkout"
/requirements-elicitation:interview "End User" --domain "checkout"

# Then discover consolidates all sources
/requirements-elicitation:discover "checkout" --sources interviews