Review fd After Customer Answers.prompt

# 72 - Review FD After Customer Answers

## Purpose

Review the customer-facing Functional Design generated after customer/domain expert answers and its internal traceability before final human/domain expert review.

This review verifies that:

1. The FD after Q&A is customer-facing clean.
2. Customer/domain expert answers were applied correctly and not over-applied.
3. Original source evidence and supplemental answers are represented separately and safely.
4. Unresolved items remain as Open Questions.
5. All meaningful FD content is traceable through `output/71_FD_AFTER_QA_INTERNAL_TRACEABILITY.md`.

This task must not rewrite the FD, generate DD, add new requirements, or invent missing evidence.

## Operating rules

- Use only the source artifacts listed in this prompt.
- Do not invent requirements, field meanings, business rules, messages, screen behavior, diagrams, or implementation details.
- Customer/domain expert answers are supplemental evidence; do not pretend they were part of the original BRD.
- Keep customer-facing content free from internal pipeline terms and internal IDs.
- Internal review reports may use internal IDs.
- Mark ambiguous, missing, contradictory, or partially supported content instead of guessing.
- Keep all output in English.

### Strikethrough / deprecated evidence rule

- Strikethrough-only, deleted, deprecated, superseded, or draft-removed source evidence must not be treated as active FD behavior unless the source or customer/domain expert answer explicitly confirms it.
- If the FD after Q&A uses strikethrough-only evidence as active behavior without confirmation, treat it as a Major or Critical issue depending on impact.

## Tasks

### Precondition

Use these latest after-QA artifacts:

- `output/70_FD_DRAFT_AFTER_QA.md` as the customer-facing FD under review.
- `output/71_FD_AFTER_QA_INTERNAL_TRACEABILITY.md` as the internal traceability evidence.
- `output/63_requirement_inventory_after_qa.md`, `output/64_business_rule_catalog_after_qa.md`, and `output/65_open_questions_after_qa.md` as the after-QA source-of-truth artifacts.
- `output/61_customer_answer_inventory.md` and `output/62_customer_answer_impact_analysis.md` as supplemental answer evidence.

Stop if `output/70_FD_DRAFT_AFTER_QA.md` or `output/71_FD_AFTER_QA_INTERNAL_TRACEABILITY.md` is missing.

### Review scope

Review these layers separately:

#### Layer 1: Customer answer application

Check whether customer/domain expert answers were applied correctly:

- Answers are applied only to relevant FD content.
- Answers are not over-applied to unrelated requirements/rules.
- Partial answers remain partially open.
- Unclear answers remain Open Questions.
- Contradictions are not hidden.
- Supplemental answers are not presented as original BRD text.

#### Layer 2: Customer-facing cleanliness

Check `output/70_FD_DRAFT_AFTER_QA.md` for:

- Professional Functional Design structure
- Clear English and technical readability
- No visible internal pipeline artifacts
- No exposed internal IDs
- No references to internal files
- No mentions of requirement inventory, business rule catalog, glossary, image analysis, translation review, impact analysis, traceability, prompts, AI, or intermediate artifacts
- Proper use of `TBD` and `Requires confirmation`
- Proper treatment of unsupported behavior
- Proper treatment of figure-derived behavior
- Proper handling of customer-answer-derived behavior
- Proper Markdown image references and Mermaid diagrams

#### Layer 3: Evidence and traceability quality

Check `output/71_FD_AFTER_QA_INTERNAL_TRACEABILITY.md` against:

- `output/63_requirement_inventory_after_qa.md`
- `output/64_business_rule_catalog_after_qa.md`
- `output/65_open_questions_after_qa.md`
- `output/61_customer_answer_inventory.md`
- `output/62_customer_answer_impact_analysis.md`
- `output/21_glossary.md`
- `output/20_image_analysis.md`
- `output/12_normalized_evidence.md`

Find:

- FD statements without traceability
- FD statements partially traced
- FD statements stronger than source evidence or customer answer evidence
- FD statements created from assumptions
- Customer answers incorrectly applied
- Customer answers missing from the FD when they should be reflected
- Original source contradictions hidden or incorrectly resolved
- Business rules missing from FD
- Unsupported cases missing from FD
- Open questions missing from FD
- Glossary-controlled terms misused in FD

#### Layer 4: Visual evidence quality

Check visual content in `output/70_FD_DRAFT_AFTER_QA.md`:

- Markdown image references
- Mermaid flowcharts
- Mermaid sequence diagrams
- Mermaid state diagrams
- Visual captions and visual notes

Find:

- Missing or broken image paths
- Decorative images without requirement value
- Mermaid diagrams with unsupported nodes, states, actors, timing, or transitions
- Figure-derived behavior overstated as confirmed text/customer-confirmed behavior
- Visuals that duplicate text without adding clarity
- Relevant source diagrams omitted without reason
- Diagram ambiguity not carried into Open Questions

## Inputs

### Primary inputs

- `output/70_FD_DRAFT_AFTER_QA.md`
- `output/71_FD_AFTER_QA_INTERNAL_TRACEABILITY.md`
- `output/63_requirement_inventory_after_qa.md`
- `output/64_business_rule_catalog_after_qa.md`
- `output/65_open_questions_after_qa.md`
- `output/61_customer_answer_inventory.md`
- `output/62_customer_answer_impact_analysis.md`
- `output/60_CUSTOMER_QA.md`
- `output/21_glossary.md`

### Supporting inputs, open only when needed

- `output/20_image_analysis.md`
- `output/12_normalized_evidence.md`
- `output/11_translation_policy.md`
- `output/10_document_inventory.md`
- `working/extracted/document_text.md`
- `working/extracted/tables.md`
- `working/extracted/image_inventory_raw.md`

## Outputs

### Output file to create or update

- `output/72_FD_AFTER_QA_REVIEW_REPORT.md`

## Required output structure

Write `output/72_FD_AFTER_QA_REVIEW_REPORT.md` using this structure: