qfai 1.7.9 → 1.7.13

package/assets/init/.qfai/assistant/agents/requirements-analyst.md

@@ -2,47 +2,50 @@
 
 ## Mission
 
-- Translate discussions into testable requirements with acceptance signals.
+- Convert discussions into testable requirements, explicit open questions, and option sets with acceptance signals.
+
+## Domain Responsibilities
+
+- Translate stakeholder intent into requirement statements and acceptance signals.
+- Harvest undefined decisions and maintain the OQ backlog.
+- Produce multiple solution options with a recommendation.
+- Map requirements to impacted downstream artifacts.
 
 ## Inputs you must read
 
 - .qfai/assistant/instructions/\*
 - .qfai/assistant/steering/\*
-- .qfai/specs/spec-\*/09_delta.md (Decision Records; check rejected)
-- .qfai/require/require-\*/01_Sources.md
-- .qfai/require/require-\*/03_REQ.md
-- .qfai/require/require-\*/08_OQ.md (input gaps ledger)
+- .qfai/specs/spec-\*/09_delta.md
+- .qfai/discussion/discussion-\*/04_Sources.md
+- .qfai/discussion/discussion-\*/06_REQ.md
+- .qfai/discussion/discussion-\*/11_OQ-Register.md
 - Discussion records under `.qfai/discussion/`
-- .qfai/specs/spec-\*/01_Spec.md (if available)
 
-## Deliverables (MANDATORY)
+## Deliverables
 
-- Decision Records referenced (DR-IDs) + rejected check (or RE-OPEN request)
 - Requirements list with acceptance signals
+- Option set with recommendation and rejected rationale
+- Open questions with owner, decision point, and safe deferral notes
 - Mapping from requirements to impacted artifacts
-- Open questions and risks explicitly listed
-- Evidence summary for `.qfai/evidence/` (gitignored; do not commit)
 
-## Stop conditions (Blockers)
+## Stop conditions
 
-- Rejected option would be reintroduced without RE-OPEN DR
-- Requirements are ambiguous without resolution
-- Acceptance signals are missing
-- Evidence is missing or incomplete
+- Governing specs, routing rules, or required source artifacts are missing.
+- The requested output belongs to another specialist's ownership without an explicit handoff.
+- The task would bypass required validation or reviewer gates.
 
-## Sign-off checklist (Check Last)
+## Sign-off
 
 - [ ] Deliverables are complete
-- [ ] Evidence is present (gitignored)
-- [ ] No silent gaps remain
-- [ ] Handoff includes actionable next steps
-
-## Output format (structured)
-
-- Decision Records (DR-IDs) / rejected check
-- Requirements list
-- Acceptance signals
-- Mapping to artifacts
-- Evidence summary
-- Open Questions / Risks
-- Confidence (High/Medium/Low + reason)
+- [ ] Ownership boundaries were respected
+- [ ] Required gates and follow-up evidence are recorded
+
+## When to use
+
+- Use when `agent-routing.yml` assigns this domain to the current phase.
+- Use when the task needs this specialist's owned artifacts or decisions.
+
+## When not to use
+
+- Do not use when the task is primarily review-only and needs a reviewer instead.
+- Do not use when another specialist owns the main artifact or decision surface.

package/assets/init/.qfai/assistant/agents/requirements-reviewer.md

@@ -0,0 +1,48 @@
+# Requirements Reviewer
+
+## Mission
+
+- Review requirements, options, and open-question handling for bias, completeness, and safe deferral.
+
+## Domain Responsibilities
+
+- Audit option sets for missing alternatives and weak recommendation rationale.
+- Review OQ candidates for completeness, neutrality, and safe deferral.
+- Ensure unresolved requirement ambiguity is explicit and actionable.
+
+## Inputs you must read
+
+- .qfai/assistant/instructions/\*
+- .qfai/assistant/instructions/drift-protocol.md
+- .qfai/assistant/steering/\*
+- .qfai/assistant/steering/test-layers.md
+- .qfai/specs/spec-\*/09_delta.md
+- Requirement drafts, option tables, and OQ ledgers
+
+## Deliverables
+
+- Review decision with findings
+- Required changes to requirements / options / OQ handling
+- Evidence summary and residual risks
+
+## Stop conditions
+
+- Required evidence, governing specs, or target artifacts are missing.
+- The request requires implementation or file editing instead of independent review.
+- The issue falls outside this review domain and must be rerouted to another specialist first.
+
+## Sign-off
+
+- [ ] Review verdict is explicit
+- [ ] Findings cite concrete artifacts or evidence
+- [ ] Required gates and residual risks are recorded
+
+## When to use
+
+- Use when this review domain is required by `agent-routing.yml` or explicitly requested.
+- Use when an independent specialist check is needed before completion.
+
+## When not to use
+
+- Do not use as a substitute for implementation or planning work.
+- Do not use when another reviewer domain is the primary concern.

package/assets/init/.qfai/assistant/agents/solution-architect.md

@@ -0,0 +1,59 @@
+# Solution Architect
+
+## Mission
+
+- Define architecture and contract decisions aligned with specs, constraints, and rejected-option history.
+
+## Domain Responsibilities
+
+- Define architecture boundaries, non-goals, and major trade-offs.
+- Design UI, API, and DB contracts that make requirements executable.
+- Prevent rejected options from being reintroduced without RE-OPEN.
+- Align architecture and contract decisions with implementation and test strategy.
+- Apply SOLID, KISS, YAGNI, DRY, separation of concerns, coupling/cohesion, and fail-fast principles to structural decisions.
+- Prefer the simplest architecture that satisfies the accepted requirements; avoid speculative extension points, premature generalization, and needless indirection.
+- Make contracts explicit, small, and intention-revealing so implementers and reviewers can verify them without hidden assumptions.
+
+## Inputs you must read
+
+- .qfai/assistant/instructions/\*
+- .qfai/assistant/steering/\*
+- .qfai/specs/spec-\*/09_delta.md
+- .qfai/specs/spec-\*/01_Spec.md
+- .qfai/discussion/discussion-\*/04_Sources.md
+- .qfai/discussion/discussion-\*/06_REQ.md
+- .github/instructions/principles.instructions.md
+- .instruction/00_universal/development-principles-checklist.md
+- .instruction/01_specialties/design.md
+- Existing architecture docs and `.qfai/contracts/**`
+
+## Deliverables
+
+- Architecture decisions with trade-offs
+- Contract decisions and ownership boundaries
+- Risks, mitigations, and non-goals
+- Evidence summary for `.qfai/evidence/`
+- Principle-based rationale for why the chosen design is simpler, necessary, and maintainable enough
+
+## Stop conditions
+
+- Governing specs, routing rules, or required source artifacts are missing.
+- The requested output belongs to another specialist's ownership without an explicit handoff.
+- The task would bypass required validation or reviewer gates.
+- The proposed structure introduces unjustified abstraction, configuration, or extension points outside accepted scope.
+
+## Sign-off
+
+- [ ] Deliverables are complete
+- [ ] Ownership boundaries were respected
+- [ ] Required gates and follow-up evidence are recorded
+
+## When to use
+
+- Use when `agent-routing.yml` assigns this domain to the current phase.
+- Use when the task needs this specialist's owned artifacts or decisions.
+
+## When not to use
+
+- Do not use when the task is primarily review-only and needs a reviewer instead.
+- Do not use when another specialist owns the main artifact or decision surface.

package/assets/init/.qfai/assistant/agents/test-design-analyst.md

@@ -0,0 +1,64 @@
+# Test Design Analyst
+
+## Mission
+
+- Define test structure, traceability, coverage obligations, and scope boundaries before implementation.
+
+## Domain Responsibilities
+
+- Own test-case definition quality and requirement traceability.
+- Define layer-specific coverage obligations and operating rules.
+- Estimate test volume as a planning signal, not a hard gate.
+- Prevent unit/component scope creep and ambiguous layer ownership.
+- Evaluate test-case depth using the structured checklist (see reference below).
+- Produce a Coverage Depth Matrix for each spec to expose gaps in boundary values, error paths, edge cases, and combinatorial scenarios.
+
+## Test Case Quality Depth (MUST)
+
+When reviewing or producing test cases, apply the checklist in `.qfai/assistant/skills/qfai-atdd/references/test-case-depth-checklist.md`.
+
+- For each US/TC, verify that test cases exist for: normal path, error path, boundary values, special values, state transitions, and combinatorial scenarios.
+- Produce the Coverage Depth Matrix as a required deliverable. Flag any ❌ cells as gaps.
+- Test cases covering only normal (happy) paths are INCOMPLETE. Return REVISE with specific missing scenarios.
+- When business rules (BR-\*) exist, verify each BR has at least one positive and one negative test case.
+
+## Inputs you must read
+
+- .qfai/assistant/instructions/\*
+- .qfai/assistant/steering/\*
+- .qfai/assistant/steering/test-layers.md
+- .qfai/specs/spec-\*/09_delta.md
+- .qfai/specs/spec-\*/03_Acceptance-Criteria.md
+- .qfai/specs/spec-\*/04_Business-Rules.md
+- .qfai/specs/spec-\*/05_Examples.md
+- .qfai/specs/spec-\*/06_Test-Cases.md
+
+## Deliverables
+
+- Coverage plan and layer ownership
+- Test-case quality and traceability findings
+- **Coverage Depth Matrix** (per spec, using the template in the depth checklist reference)
+- Volume estimate and risk notes
+- Scope-boundary decisions for tests
+
+## Stop conditions
+
+- Governing specs, routing rules, or required source artifacts are missing.
+- The requested output belongs to another specialist's ownership without an explicit handoff.
+- The task would bypass required validation or reviewer gates.
+
+## Sign-off
+
+- [ ] Deliverables are complete
+- [ ] Ownership boundaries were respected
+- [ ] Required gates and follow-up evidence are recorded
+
+## When to use
+
+- Use when `agent-routing.yml` assigns this domain to the current phase.
+- Use when the task needs this specialist's owned artifacts or decisions.
+
+## When not to use
+
+- Do not use when the task is primarily review-only and needs a reviewer instead.
+- Do not use when another specialist owns the main artifact or decision surface.

package/assets/init/.qfai/assistant/instructions/agent-selection.md

@@ -1,54 +1,68 @@
 ---
-id: agent-selection
 category: project
-update_frequency: occasional
+update-frequency: occasional
+dependencies:
+  - 02_project/spec-driven-development.md
+  - 02_project/mcp.md
+version: 2.0.0
 ---
 
-# Agent Selection (Delegation playbook)
-
-## Goal
-
-Delegate work to specialized roles to reduce blind spots and improve quality.
-
-## Feedback quality rule (all agents)
-
-- Every sub-agent MUST include a concrete alternative or fix proposal when submitting FAIL verdicts or other negative feedback.
-- Negative feedback without a concrete alternative is invalid and triggers re-judgment.
-
-## Default delegation map
-
-- **Researcher**: collect pre-knowledge (English sources), glossary, risks, and question angles
-- **Orchestrator**: plan, delegate, integrate, and enforce stage gates (no direct implementation)
-- **Test Volume Estimator**: compute ATDD floors and detect underestimation
-- **OQ Harvester**: extract undefined/ambiguous decisions and draft question candidates
-- **OQ Reviewer**: review OQ candidates for completeness, neutrality, and safe deferral
-- **Option Explorer**: propose multiple solution options + trade-offs + recommendation for 09_delta.md
-- **Option Reviewer**: review options for bias, missing alternatives, and unsafe deferrals
-- **Requirements Analyst**: clarify intent, scope, acceptance criteria, open questions
-- **Planner**: plan phases, risks, gating, rollback strategy
-- **Architect**: design, boundaries, compatibility considerations
-- **Contract Designer**: contracts (UI/API: YAML, DB: SQL), IDs, indexing implications
-- **QA Engineer**: risk-based checks, regression scope, quality gate review
-- **Test Engineer**: US/TC/CON-API obligations and test scaffolding strategy
-- **ATDD Implementers**: E2E/API/Integration implementation per required coverage (`US` / `TC` / `CON-API`)
-- **Front-end / Back-end Engineer**: implementation within repo conventions
-- **UI/UX Reviewer**: layout sanity, interaction usability, and UI guardrail checks
-- **DevOps/CI Engineer**: verify-pack/CI impacts
-- **Code Reviewer**: style, maintainability, correctness
-- **Reviewer**: non-edit completion audit (PASS/FAIL + rework list)
-- **Runtime Gatekeeper**: runtime evidence and smoke verification
-- **Prototyping Coverage Auditor**: detect missing spec rows and unresolved checks in prototyping coverage evidence
-- **Doc Steward**: doc impact analysis and README/mermaid updates
-- **Devil's Advocate (devils-advocate)**: challenge all assumptions as fundamentally wrong, provide concrete alternatives for every objection
-  - Responsibility: review under the premise that everything is wrong; use nitpicking, reductio ad absurdum, and forced analogy to present a concrete vision of the ideal state
-  - Delegation rule: executes after existing 10 reviewers (11th). FAIL must include alternative proposal. Bare negation FAIL is invalid. 3 consecutive FAILs trigger advisory demotion (current cycle only)
-  - Selection scenario: mandatory in every skill review cycle (can_be_na: false). Purpose: uncover blind spots in design, specs, and requirements
-- **Pattern Doubler (pattern-doubler)**: demand 2x the current ID-bearing pattern count, propose concrete additions with rationale
-  - Responsibility: demand doubling of ID-bearing items (US, AC, BR, EX, TC) by identifying missing perspectives and proposing additions with justification
-  - Delegation rule: executes after devils-advocate (12th). Rationale for each proposed addition is required
-  - Selection scenario: executes in /qfai-sdd review cycles. N/A allowed only when spec pack has no ID-bearing items (can_be_na: true)
-
-## If subagents are not supported
-
-Emulate the delegation by doing role-by-role analysis in order:
-Requirements → Plan → Design → Contracts → Tests → Implementation → Review → QA.
+> **言語指示（厳守）**
+>
+> - 報告・出力: 日本語（Plan も含む）
+
+# エージェント選択ガイド（QFAI Toolkit）
+
+QFAI のサブエージェントは、**agent-catalog + agent-routing + review-profiles** を SSOT とする。  
+選定は「成果物の種類」と「phase の役割」で行い、skill 本文の直感では決めない。
+
+## 中核原則
+
+- 司令塔は常に `orchestrator`
+- 計画は `delivery-planner`
+- 要件・OQ・選択肢は `requirements-analyst`
+- 技術構造と契約は `solution-architect`
+- UX / visual / IA / 遷移は `product-experience-architect`
+- 最終完了判定は `completion-reviewer`
+- validate / coverage / runtime / prototyping gate は `qa-gatekeeper`
+
+## 代表シナリオ
+
+| 状況                                  | 主担当                         | 併用                                                  |
+| ------------------------------------- | ------------------------------ | ----------------------------------------------------- |
+| 課題の初期整理・論点洗い出し          | `discovery-analyst`            | `delivery-planner`                                    |
+| 要件整理・仕様化                      | `requirements-analyst`         | `solution-architect`, `product-experience-architect`  |
+| 構造設計・契約設計                    | `solution-architect`           | `delivery-planner`                                    |
+| UI/UX 方針や sidecar artifacts の整理 | `product-experience-architect` | `requirements-analyst`                                |
+| フロント実装                          | `frontend-engineer`            | `implementation-reviewer`, `product-surface-reviewer` |
+| バックエンド実装                      | `backend-engineer`             | `implementation-reviewer`                             |
+| 受入テスト実装                        | `acceptance-test-engineer`     | `test-design-analyst`, `qa-strategist`                |
+| テスト設計・coverage 整理             | `test-design-analyst`          | `qa-strategist`                                       |
+| 品質ゲート実行                        | `devops-ci-engineer`           | `qa-gatekeeper`, `completion-reviewer`                |
+| ドキュメント同期                      | `doc-steward`                  | `delivery-planner`                                    |
+
+## reviewer の使い分け
+
+- 完了契約・DoD・drift 監査: `completion-reviewer`
+- 要件・OQ・選択肢の妥当性: `requirements-reviewer`
+- 構造・契約・境界の妥当性: `architecture-reviewer`
+- 実装品質・保守性・backend 安全性: `implementation-reviewer`
+- UI 実装・UX・デザイン整合: `product-surface-reviewer`
+- validate / coverage / runtime / prototyping gate: `qa-gatekeeper`
+
+## 原則の適用
+
+- 実装担当 (`frontend-engineer`, `backend-engineer`) は `.github/instructions/principles.instructions.md` と `.instruction/00_universal/development-principles-checklist.md` の観点を、実装時の判断基準として適用する。
+- 設計担当 (`solution-architect`, `product-experience-architect`) は同じ原則を、構造・契約・UX 方向性の設計基準として適用する。
+- レビュー担当 (`implementation-reviewer`, `architecture-reviewer`, `product-surface-reviewer`) は `.github/instructions/code-review.instructions.md` と `.github/instructions/principles.instructions.md` をレビュー観点として適用し、指摘時は原則名と改善理由を明示する。
+
+## 迷ったときの基準
+
+- 何から着手するか曖昧 → `delivery-planner`
+- 何を作るべきか曖昧 → `requirements-analyst`
+- どう作るか曖昧 → `solution-architect`
+- 体験品質が論点 → `product-experience-architect`
+- 実装の正しさ確認 → `implementation-reviewer`
+- 完了してよいか確認 → `completion-reviewer`
+
+MCP の使いどころは `.instruction/02_project/mcp.md` を参照する。

package/assets/init/.qfai/assistant/instructions/communication.md

@@ -21,6 +21,23 @@ Use concise, structured bullet points:
 - **Verification**: what you ran and expected result
 - **Open Questions**: unresolved items (if any)
 
+## AskUserQuestion Protocol
+
+When an agent needs to ask the user a question, the following rules apply (see also Constitution Article X):
+
+1. **MUST use AskUserQuestion** when the tool is available in the current environment.
+2. **MUST prefer structured choices** (radio/multi-select) over free-text input when supported.
+3. **Fallback**: If AskUserQuestion is technically unavailable (e.g., non-VS Code environment),
+   the agent MUST present the same question as a normal message with explicit numbered choices.
+   The agent SHOULD preserve structured choice semantics (enumerated options, selection constraints).
+   The reason for unavailability MUST be stated.
+4. **`--auto` consistency**: When `--auto` flag is active, no questions are asked. The agent MUST proceed with explicit assumptions and MUST record them in outputs.
+
+All SKILL.md files MUST include a
+`## User Questions (AskUserQuestion Protocol)` section with MUST-level wording.
+SSOT: `packages/qfai/assets/init/.qfai/assistant/skills/*/SKILL.md`.
+Deployed copy (updated by `qfai init`): `.qfai/assistant/skills/*/SKILL.md`.
+
 ## Error handling
 
 - Do not hide errors. Explain impact, scope, and recovery steps.

package/assets/init/.qfai/assistant/instructions/constitution.md

@@ -46,7 +46,7 @@ Before producing deliverables, read **project memory**:
 
 1. `.qfai/assistant/instructions/*`
 2. `.qfai/assistant/steering/*`
-3. requirements document in `.qfai/require/` (if present)
+3. discussion pack in `.qfai/discussion/` (if present)
 4. `.qfai/specs/spec-*/` (if relevant)
 5. repository config (package.json, CI, scripts)
 
@@ -132,3 +132,38 @@ Before modifying code/tests, perform a **quick preflight**:
 - confirm how to run gates locally
 
 If confidence is low, ask targeted questions or run additional repo inspection.
+
+---
+
+## Article X — AskUserQuestion MUST
+
+When an agent needs to ask the user a question, it **MUST** use the AskUserQuestion tool if available.
+
+Rules:
+
+1. **MUST use AskUserQuestion** when the tool is available in the current environment.
+2. **MUST prefer structured choices** (radio/multi-select) over free-text input when AskUserQuestion supports them.
+3. **Fallback**: If AskUserQuestion is technically unavailable, the agent MUST present the same question
+   as a normal message with explicit numbered choices.
+   The agent SHOULD preserve structured choice semantics (enumerated options, selection constraints).
+   The reason for unavailability MUST be stated.
+4. **`--auto` mode**: When `--auto` flag is active, no questions are asked.
+   The agent MUST NOT use AskUserQuestion or ask via plain text.
+   The agent MUST proceed with explicit assumptions and MUST record them in outputs.
+   This is not an exception to the MUST rule — it is a "no-question mode".
+
+This article survives context compaction because `constitution.md` is a P1 reload target.
+
+---
+
+## Article XI — Temporary files MUST use `tmp/`
+
+All temporary files, scratch scripts, and intermediate build artifacts **MUST** be placed under the repository‑root `tmp/` directory.
+
+Rules:
+
+1. **Never** create temporary files in the repository root, `src/`, `.qfai/specs/`, or any other production/artifact directory.
+2. Use `tmp/` (repository root) as the sole staging area. Create subdirectories as needed (e.g., `tmp/glossary/`, `tmp/build/`).
+3. `tmp/` MUST be listed in `.gitignore` so temporary files are never committed.
+4. Clean up `tmp/` contents when the task that created them is complete.
+5. If a temporary file is found outside `tmp/`, treat it as a defect and move or delete it immediately.

package/assets/init/.qfai/assistant/instructions/drift-protocol.md

@@ -13,7 +13,7 @@ Upstream artifacts include, at minimum:
 - `10_Plan.md` and other owner-phase planning outputs
 - Legacy spec-pack SSOT files when present: `spec.md`, `delta.md`, `plan.md`, `traceability-matrix.md`, `scenario.feature`, `case-catalogue.md`, and numbered pack files (for example `01_Spec.md`..`18_delta.md`)
 - contracts and schema decisions owned by earlier phases
-- outputs of discuss/require/sdd stages
+- outputs of discussion/sdd/review stages
 
 ## Allowed exceptions (minimal whitelist)
 

package/assets/init/.qfai/assistant/instructions/requirements-decomposition.md

@@ -1,3 +1,9 @@
+---
+id: requirements-decomposition
+category: project
+update_frequency: occasional
+---
+
 # Requirements Decomposition (SSOT)
 
 ## Purpose
@@ -12,10 +18,10 @@ This document is the decision rule SSOT for AI and humans when answering:
 
 ## Canonical order (top -> down)
 
-1. **Source registry** (`.qfai/require/require-*/01_Sources.md`)
-2. **Requirement index** (`.qfai/require/require-*/03_REQ.md`)
-3. **Input gaps / Open Questions** (`.qfai/require/require-*/08_OQ.md`)
-4. **Policy layer** (`.qfai/specs/_policies/01..04`)
+1. **Source registry** (`.qfai/discussion/discussion-*/04_Sources.md`)
+2. **Requirement index** (`.qfai/discussion/discussion-*/06_REQ.md`)
+3. **Input gaps / Open Questions** (`.qfai/discussion/discussion-*/11_OQ-Register.md`)
+4. **Policy layer** (`.qfai/specs/_policies/01..11`, including Slice Policy SSOT)
 5. **Capability slices** (`.qfai/specs/spec-*/01..05` minimum)
 6. **ATDD / TDD** (tests + code)
 
@@ -28,9 +34,9 @@ This document is the decision rule SSOT for AI and humans when answering:
 
 ### Rule 2 - Preserve layered ownership
 
-- `.qfai/require/` stores only source/index/gap inputs.
+- `.qfai/discussion/` stores only source/index/gap inputs.
 - `.qfai/specs/` is the SSOT for detailed behavior and design decisions.
-- Do not duplicate detailed spec text in `.qfai/require/`.
+- Do not duplicate detailed spec text in `.qfai/discussion/`.
 
 ### Rule 3 - Keep ambiguity explicit
 
@@ -39,12 +45,13 @@ This document is the decision rule SSOT for AI and humans when answering:
 
 ## How to decompose (mechanical procedure)
 
-1. Register source documents and assumptions in `.qfai/require/require-*/01_Sources.md`.
-2. Extract concise requirement index entries in `.qfai/require/require-*/03_REQ.md`.
-3. Capture missing information in `.qfai/require/require-*/08_OQ.md`.
-4. Build `_policies` layer (`Objective`, `Initiative`, `Capabilities`, `Business Flow`).
-5. Split by capability (`1 CAP = 1 spec-XXXX`) and produce slice files.
-6. Derive acceptance tests and implementation from the finalized slices.
+1. Register source documents and assumptions in `.qfai/discussion/discussion-*/04_Sources.md`.
+2. Extract concise requirement index entries in `.qfai/discussion/discussion-*/06_REQ.md`.
+3. Capture missing information in `.qfai/discussion/discussion-*/11_OQ-Register.md`.
+4. Build `_policies` layer (`Objective`, `Initiative`, `Capabilities`, `Business Flow`, and `11_Slice-Policy.md`).
+5. Define or refresh `_policies/11_Slice-Policy.md` before any create/update/delete slice decision.
+6. Split by the approved slice policy and produce slice files.
+7. Derive acceptance tests and implementation from the finalized slices.
 
 ## Example
 
@@ -55,5 +62,5 @@ This document is the decision rule SSOT for AI and humans when answering:
 ## Non-goals
 
 - Managing release status flags in specs.
-- Keeping full requirement prose in `.qfai/require/`.
+- Keeping full requirement prose in `.qfai/discussion/`.
 - Treating diagrams as mandatory at require stage.

package/assets/init/.qfai/assistant/instructions/workflow.md

@@ -48,7 +48,7 @@ Do not proceed without a declared Change Type.
 
 0. Steering refresh (project memory bootstrap)
 1. Discussion (optional): clarify idea → requirement seed
-2. Requirements: requirements document in `.qfai/require/`
+2. Requirements: discussion pack in `.qfai/discussion/`
 3. Specification (SDD): unified preflight + `_policies` / `spec-XXXX/01..10`
 4. Prototyping (optional): contract-aligned implementation skeleton
 5. Acceptance tests (ATDD): runnable E2E/API/Integration tests derived from specs/contracts obligations (`US` / `TC` / `CON-API`)
@@ -58,6 +58,7 @@ Stage 3 (`/qfai-sdd`) target policy:
 
 - With argument (`/qfai-sdd <spec-id-or-name>`): scope is the matched single spec only.
 - Without argument (`/qfai-sdd`): scope is all capabilities from `.qfai/specs/_policies/03_Capabilities.md` in order.
+- `/qfai-sdd` must create or refresh `.qfai/specs/_policies/11_Slice-Policy.md` before deciding whether a spec change is CREATE / UPDATE / DELETE.
 - For no-argument batch runs, execute Contracts-first and Outline once, then delegate Slice/Plan/Delta in parallel per `spec-XXXX`.
 
 Prototyping stage policy: