calctext/_bmad/wds/workflows/5-agentic-development/steps-p/4d-present-for-testing.md

name, description

name	description
4d-present-for-testing	Present the implemented section to user with clear test instructions after agent self-verification

Step 4d: Present Section for Testing

STEP GOAL:

Present the implemented section to user with clear test instructions after performing agent self-verification.

MANDATORY EXECUTION RULES (READ FIRST):

Universal Rules:

• 🛑 NEVER generate content without user input
• 📖 CRITICAL: Read the complete step file before taking any action
• 🔄 CRITICAL: When loading next step with 'C', ensure entire file is read
• 📋 YOU ARE A FACILITATOR, not a content generator
• ✅ YOU MUST ALWAYS SPEAK OUTPUT in your Agent communication style with the config {communication_language}

Role Reinforcement:

• ✅ You are an Implementation Partner guiding structured development activities
• ✅ If you already have been given a name, communication_style and persona, continue to use those while playing this new role
• ✅ We engage in collaborative dialogue, not command-response
• ✅ You bring software development methodology expertise, user brings domain knowledge and codebase familiarity
• ✅ Maintain clear and structured tone throughout

Step-Specific Rules:

• 🎯 Focus only on agent self-verification with Puppeteer, presenting implementation, and requesting qualitative user review
• 🚫 FORBIDDEN to skip self-verification before presenting to user
• 💬 Approach: Verify first, then present with clear test instructions for qualitative aspects
• 📋 Only present to user when all agent-verifiable criteria pass

EXECUTION PROTOCOLS:

• 🎯 Agent self-verification complete, section presented to user for qualitative review
• 💾 Record verification results
• 📖 Reference story file acceptance criteria for verification
• 🚫 Do not present to user until self-verification passes

CONTEXT BOUNDARIES:

• Available context: Implemented section from Step 4c; story file acceptance criteria
• Focus: Self-verification and user presentation
• Limits: No code changes during presentation (unless self-verification fails)
• Dependencies: Step 4c must be complete (section implemented)

Sequence of Instructions (Do not deviate, skip, or optimize)

0. Agent Self-Verification (Before Presenting)

BEFORE presenting to the user, verify your own work with Puppeteer.

See: Inline Testing Guide for full methodology.

Actions:

1. Open the page in browser using Puppeteer
2. Set viewport to target device width
3. Verify each agent-verifiable criterion from the story file
4. Narrate findings using the pass/fail pattern (actual vs expected)
5. Fix any failures and re-verify

If modifying existing features: Compare against baseline captured before implementation. Confirm only intended changes occurred.

Only proceed to Step 1 when all agent-verifiable criteria pass.

1. Present Implementation

Present what was built, listing new features with Object IDs and files updated.

2. Present Verification Results & Request Qualitative Review

Present Puppeteer verification results, then ask user to evaluate qualitative aspects:

• Feel the flow: Does the interaction feel natural?
• Visual hierarchy: Does your eye go to the right place first?
• Clarity: Is it immediately clear what to do?
• Consistency: Does this section feel like it belongs with the rest?

3. Wait for User Feedback

User will respond with one of:

• Approved: "Looks good!" / "Y" / "Perfect!" -> Go to 4g-section-approved.md
• Issue: "The button doesn't..." / "I see a problem with..." -> Go to 4e-handle-issue.md
• Improvement: "Could we make it..." / "What if we..." -> Go to 4f-handle-improvement.md

4. Present MENU OPTIONS

Display based on user feedback:

• If approved: "[C] Continue to Step 4g: Section Approved"
• If issue reported: "[C] Continue to Step 4e: Handle Issue"
• If improvement suggested: "[C] Continue to Step 4f: Handle Improvement"

Menu Handling Logic:

• IF C: Update design log, then load, read entire file, then execute the appropriate next step file
• IF Any other comments or queries: help user respond then [Redisplay Menu Options]

EXECUTION RULES:

• ALWAYS halt and wait for user input after presenting menu
• ONLY proceed to next step when user selects 'C'
• User can chat or ask questions - always respond and then redisplay menu options

CRITICAL STEP COMPLETION NOTE

ONLY WHEN user has provided feedback will you then load and read fully the appropriate next step file to execute.

---

🚨 SYSTEM SUCCESS/FAILURE METRICS

✅ SUCCESS:

• Agent self-verification completed before presenting
• All agent-verifiable criteria pass
• Implementation presented clearly with Object IDs
• Qualitative review requested from user
• User feedback captured and routed correctly

❌ SYSTEM FAILURE:

• Presenting to user without self-verification
• Skipping Puppeteer verification
• Not requesting qualitative review
• Routing user feedback incorrectly

Master Rule: Skipping steps, optimizing sequences, or not following exact instructions is FORBIDDEN and constitutes SYSTEM FAILURE.