qfai 1.9.2 → 1.10.0

package/assets/init/.qfai/assistant/manifest/agent-routing.yml

@@ -48,34 +48,70 @@ routing:
         rerun_policy: changed-scope-dependents
     review_profile: default
 
+  # `parallel_groups` describes ROLE FAN-OUT within a phase, not item
+  # dispatch. An empty list means no roles run concurrently; it says
+  # nothing about item-level parallelism, which `qfai-implement/SKILL.md`
+  # `## Parallelization Policy` governs.
+  # `iteration` says how often a phase runs. `qfai-implement` drives the TDD
+  # micro-cycle one ledger row at a time, so `red`, `build` and `test` run once
+  # per row. Without this key the whole phase list read as a single pass over
+  # the batch, and the only slot `qa-gatekeeper` had was the terminal `review`
+  # phase — by which point production code exists and the RED state the
+  # gatekeeper is supposed to observe has been destroyed. The gate could then
+  # adjudicate nothing but the implementer's own written account, which is the
+  # self-attestation it exists to prevent.
   - skill: qfai-implement
     phases:
       - id: plan
+        iteration: per-invocation
         mandatory_agents: [delivery-planner, test-design-analyst]
         conditional_agents: []
         parallel_groups: []
         blocking_agents: [delivery-planner]
         rerun_policy: failed-agents-only
+      # The failing test exists and no production code does. This is the only
+      # window in which "watch it fail" is a first-hand observation, so the
+      # gatekeeper is mandatory AND blocking here: RED must be confirmed before
+      # the row is allowed to proceed to `build`.
+      - id: red
+        iteration: per-ledger-item
+        mandatory_agents: [qa-gatekeeper]
+        conditional_agents: []
+        parallel_groups: []
+        blocking_agents: [qa-gatekeeper]
+        rerun_policy: failed-agents-only
       - id: build
-        mandatory_agents: []
+        iteration: per-ledger-item
+        mandatory_agents: [qa-gatekeeper]
         conditional_agents: [frontend-engineer, backend-engineer]
         parallel_groups:
           - [frontend-engineer, backend-engineer]
-        blocking_agents: []
+        blocking_agents: [qa-gatekeeper]
         rerun_policy: changed-scope-dependents
       - id: test
+        iteration: per-ledger-item
         mandatory_agents: []
         conditional_agents: [acceptance-test-engineer, qa-strategist]
         parallel_groups: []
         blocking_agents: []
         rerun_policy: changed-scope-dependents
+      # Still per-item: the reviewers judge one row (or one T1 coherent group),
+      # never the whole ledger at once. `qa-gatekeeper` stays listed because the
+      # completion gate re-reads its verdicts here; its RED/GREEN observations
+      # were already made in `red` and `build`.
       - id: review
+        iteration: per-ledger-item
         mandatory_agents: [implementation-reviewer, qa-gatekeeper, completion-reviewer]
         conditional_agents: [product-surface-reviewer]
         parallel_groups: []
-        blocking_agents: [qa-gatekeeper, completion-reviewer]
+        # `implementation-reviewer` is blocking, not advisory: it is item 8 of the
+        # unconditional 11-point gate in qfai-implement/SKILL.md, and every DONE
+        # rule in the skill and in shared-skill-delegation-baseline.md is phrased
+        # over *blocking* reviewers. Listing it as mandatory-only let an item
+        # reach `done` with the code-quality reviewer sitting at REVISE.
+        blocking_agents: [implementation-reviewer, qa-gatekeeper, completion-reviewer]
         rerun_policy: changed-scope-dependents
-    review_profile: runtime-heavy
+    review_profile: implementation-heavy
 
   - skill: qfai-atdd
     phases:
@@ -176,3 +212,13 @@ routing:
         blocking_agents: [qa-gatekeeper, completion-reviewer]
         rerun_policy: changed-scope-dependents
     review_profile: runtime-heavy
+
+# `rerun_policy` vocabulary (validated by `QFAI-AGENT-013`):
+#
+# - `failed-agents-only`        re-run only the agents that did not return PASS
+# - `changed-scope-dependents`  re-run every agent whose inputs the change
+#                               touched, including ones that passed
+#
+# The key was previously on all 23 phases, defined nowhere and read by nothing,
+# so a typo in it was invisible. `constitution/drift-protocol.md` step 4 uses the
+# same vocabulary when a Change Request names its rerun mode.

package/assets/init/.qfai/assistant/manifest/review-profiles.yml

@@ -16,6 +16,15 @@ profiles:
   runtime-heavy:
     always_required: [completion-reviewer, qa-gatekeeper]
     conditional_required: [implementation-reviewer]
+  # `qfai-implement` makes `implementation-reviewer` PASS item 8 of its
+  # unconditional 11-point gate, so for that skill the code-quality review is
+  # always required and always blocking. It shares `runtime-heavy` with
+  # qfai-atdd / qfai-configure / qfai-verify, where the same reviewer is
+  # genuinely conditional — hence a profile of its own rather than widening
+  # runtime-heavy for all four.
+  implementation-heavy:
+    always_required: [completion-reviewer, qa-gatekeeper, implementation-reviewer]
+    conditional_required: []
 
 optional_modes:
   devils-advocate:

package/assets/init/.qfai/assistant/process/migrations/v1.4.27-atdd-alignment.md

@@ -27,7 +27,7 @@
    - `tests/integration`
 4. Remove forbidden references (`TC` annotations in API/E2E tests).
 5. Re-run:
-   - `qfai validate --fail-on error`
+   - `npx qfai validate --fail-on error`
 
 ## Legacy Compatibility
 

package/assets/init/.qfai/assistant/skills/qfai-atdd/SKILL.md

@@ -54,8 +54,9 @@ When unsure, read inputs in this order:
   - `.qfai/specs/<spec-id>/05_Examples.md` (EX)
   - `.qfai/specs/<spec-id>/06_Test-Cases.md` (TC)
   - `.qfai/contracts/api/**` (CON-API)
+  - `.qfai/contracts/db/**` (CON-DB)
   - `.qfai/contracts/ui/**` and `.qfai/contracts/design/**` when the target spec is UI-bearing
-- P5: `.qfai/specs/<spec-id>/09_delta.md` (Decision Records; if no spec yet, state "not applicable")
+- P5: `.qfai/specs/<spec-id>/07_Decisions.md` + `.qfai/specs/_policies/08_Decisions.md` (Decision Records, `DR-*`; if no spec yet, state "not applicable")
 - P6: legacy artifacts (optional only)
   - `.qfai/specs/<spec-id>/scenario.feature`
   - coverage ledger files
@@ -91,7 +92,7 @@ Follow `.qfai/assistant/constitution/shared-skill-delegation-baseline.md`.
 ### Delegation Failure (Hard Stop)
 
 - No additional overrides.
-- Do not simulate roles. If the first required delegation fails, stop the stage and report remediation.
+- Do not simulate roles. Classify the failure per the baseline taxonomy first: `unavailable` stops the stage with a remediation report; `saturated` uses the bounded retry branch and keeps the stage open.
 
 ## Work Orders Summary
 
@@ -111,9 +112,9 @@ Use the shared schema.
 - Follow `.qfai/assistant/constitution/shared-skill-delegation-baseline.md#reviewer-gate-baseline`.
 - Final completion gate MUST be delegated to an independent `completion-reviewer`.
 - ATDD-specific reviewer checks:
-  - coverage obligations met: E2E covers `US`, Integration covers `TC`, API covers `CON-API`;
+  - coverage obligations met: E2E covers `US`, API covers `CON-API`, and every `TC` is covered from the directory its declared `Level` routes to;
   - Coverage Depth Matrix is reviewed and no unjustified `X` cells remain;
-  - validation evidence exists and `qfai validate --profile atdd --fail-on error` passes;
+  - validation evidence exists and `npx qfai validate --profile atdd --fail-on error` passes;
   - Drift Protocol is enforced;
   - test-layer policy is checked against `.qfai/assistant/catalog/test-layers.md`;
   - coverage floors and ratios are signals, not gates;
@@ -133,7 +134,7 @@ Use the shared template.
 
 Use the shared template.
 
-- Required field: `Status (PASS/REVISE)`.
+- Required field: `Status (PASS/REVISE/PENDING)`. `PENDING` marks a gate that could not be run (see the baseline's reviewer-budget branch); it never counts as `PASS`.
 
 ## Stage 0 — Steering completion refresh (mandatory)
 
@@ -148,18 +149,28 @@ Follow `.qfai/assistant/constitution/shared-skill-operating-baseline.md#delta-re
 - Do NOT declare completion based on unit/component tests.
 - `10_Plan.md` is the primary How SSOT for execution phases.
 - If `10_Plan.md` is missing, stop and run owner planning flow before proceeding.
-- Completion gate is validation with zero errors (`qfai validate --profile atdd --fail-on error`).
+- Completion gate is validation with zero errors (`npx qfai validate --profile atdd --fail-on error`).
 - Coverage obligations are mandatory:
   - `tests/e2e/**` must cover all required `US-*`.
-  - `tests/integration/**` must cover all required `TC-*`.
+  - Every `TC-*` must be covered from the directory its declared `Level` routes
+    to: `L3`/`Integration` -> `tests/integration/**`, `L4`/`API` ->
+    `tests/api/**`, `L5`/`E2E` -> `tests/e2e/**`. A TC with no declared `Level`
+    routes to `tests/integration/**`.
   - `tests/api/**` must cover all required `CON-API-*`.
-- Forbidden references:
-  - `tests/api/**` must not contain `QFAI:SPEC-XXXX:TC-YYYY`.
-  - `tests/e2e/**` must not contain `QFAI:SPEC-XXXX:TC-YYYY`.
+- Forbidden references (a TC annotation outside its declared home):
+  - `tests/api/**` must not contain `QFAI:SPEC-XXXX:TC-YYYY` unless that TC
+    declares `Level` `L4`/`API`.
+  - `tests/e2e/**` must not contain `QFAI:SPEC-XXXX:TC-YYYY` unless that TC
+    declares `Level` `L5`/`E2E`.
 - Unknown references (`US/TC/CON-API` not declared) must be treated as errors.
 - Floors/ratios are planning signals only, not gates.
 - Legacy `scenario.feature` or coverage ledgers may exist but are not mandatory inputs for completion.
-- Evidence file is required under `.qfai/evidence/` and must not be committed.
+- Evidence file is required under `.qfai/evidence/`. Stage evidence is
+  **regenerable** and is not committed. **Governance records are different**:
+  Change Requests (`.qfai/decisions/CR-*.md`) and durable decision records
+  (`.qfai/evidence/decisions/*.json`) carry user approval, are not
+  regenerable, and stay in version control — the managed `.gitignore` block
+  negates them for that reason.
 
 ## Completion Contract (Shared)
 
@@ -168,7 +179,7 @@ Follow `.qfai/assistant/constitution/shared-skill-operating-baseline.md#gate-fai
 
 ## Goal
 
-Turn specs/contracts obligations (`US` / `TC` / `CON-API`) into runnable acceptance tests in this repository.
+Turn specs/contracts obligations (`US` / `TC` / `CON-API` / `CON-DB`) into runnable acceptance tests in this repository.
 
 ## Scope (ATDD only)
 
@@ -204,25 +215,54 @@ Turn specs/contracts obligations (`US` / `TC` / `CON-API`) into runnable accepta
 | API         |      #CON |  API_s | API contracts |       |
 | Integration |       #TC |  INT_s | test cases    |       |
 
+## Scaffolding
+
+`npx qfai atdd scaffold --spec <spec-id>` bulk-emits one placeholder test per
+coverage-target `TC-*`, each carrying its `QFAI:SPEC-XXXX:TC-YYYY` annotation.
+Skeletons land in `tests/integration/<spec-id>/` — the directory
+`QFAI-ATDD-112` scans — so a filled-in skeleton counts as coverage. It is
+idempotent: existing files are left untouched.
+
+A skeleton left in placeholder shape across repeated validate runs escalates
+(`qfai.config.yaml#atdd.scaffoldEscalateCycles`), so scaffolding is a start, not
+a discharge of the obligation.
+
 ## Annotation obligations (mandatory)
 
 Every generated ATDD test MUST include QFAI annotations by layer:
 
-- `tests/e2e/**`: `QFAI:SPEC-XXXX:US-YYYY`
-- `tests/integration/**`: `QFAI:SPEC-XXXX:TC-YYYY`
-- `tests/api/**`: `QFAI:CON-API-XXXX`
+- `tests/e2e/**`: `QFAI:SPEC-XXXX:US-YYYY` (plus `QFAI:SPEC-XXXX:TC-YYYY` for a
+  TC that declares `Level` `L5`/`E2E`)
+- `tests/integration/**`: `QFAI:SPEC-XXXX:TC-YYYY` (TCs declaring `L3`/
+  `Integration`, and TCs with no declared `Level`)
+- `tests/api/**`: `QFAI:CON-API-XXXX` (plus `QFAI:SPEC-XXXX:TC-YYYY` for a TC
+  that declares `Level` `L4`/`API`)
+- `tests/integration/**` also carries `QFAI:CON-DB-XXXX` for every declared DB
+  contract the slice exercises
 
 Notes:
 
+- A TC's annotation belongs in exactly one directory — the one its declared
+  `Level` routes to. Placing it elsewhere is both uncovered (`QFAI-ATDD-112`)
+  and forbidden (`QFAI-ATDD-121` / `QFAI-ATDD-122` / `QFAI-ATDD-123`). The rule
+  is symmetric: an annotation left behind in `tests/integration/**` after the
+  TC moved to `L4`/`L5` is rejected by `QFAI-ATDD-123` just as an early one in
+  `tests/api/**` is rejected by `QFAI-ATDD-121`.
 - AC annotations are optional in code.
 - `QFAI:CON-API-*` in E2E is not forbidden, but contract guarantee belongs to API tests.
 
 ## Success Criteria (Definition of Done)
 
 - All required `US` are covered by E2E tests.
-- All required `TC` are covered by integration tests.
+- All required `TC` are covered from the directory their declared `Level` routes
+  to (`L3`/`Integration` -> `tests/integration/**`, `L4`/`API` ->
+  `tests/api/**`, `L5`/`E2E` -> `tests/e2e/**`, no declared `Level` ->
+  `tests/integration/**`). Duplicating a TC into a second layer is a
+  not-done condition, not extra credit.
 - All required `CON-API` are covered by API tests.
-- Validation passes: `qfai validate --profile atdd --fail-on error`.
+- All required `CON-DB` are covered by integration tests (`QFAI-ATDD-115`); a contract
+  outside the current slice is deferred with `-- x-qfai-status: planned`, not left uncovered.
+- Validation passes: `npx qfai validate --profile atdd --fail-on error`.
 - Repository quality gates (format/lint/type/tests/pack) pass with evidence.
 - Evidence file exists and includes work orders + reviewer notes.
 - Completion is approved by a reviewer who did not implement tests.
@@ -304,7 +344,7 @@ Template:
 ## Stage Gates (Do not skip)
 
 - P0: Plan and obligations checklist prepared.
-- P1: Layer assignment validated.
+- P1: Layer assignment validated against `.qfai/assistant/catalog/test-layers.md#layer-derivation-procedure-normative`.
 - P2: E2E implementation completed.
 - P3: API implementation completed.
 - P4: Integration implementation completed.
@@ -321,7 +361,7 @@ Before declaring completion:
 2. Run:
 
    ```bash
-   qfai validate --profile atdd --fail-on error
+   npx qfai validate --profile atdd --fail-on error
    ```
 
 3. Run repository standard gates:
@@ -397,6 +437,6 @@ A skill MAY narrow the auto-decide bucket (drop entries) but MUST NOT widen it.
 
 project_memory:
 
-- Coverage obligations stay layer-pinned: tests/e2e/** must cover all required US; tests/integration/** all required TC; tests/api/\*\* all required CON-API.
-- Forbidden references guard the test-layer policy: tests/api/** must not carry QFAI:SPEC-XXXX:TC-YYYY annotations; tests/e2e/** likewise.
+- Coverage obligations stay layer-pinned for US and CON-API: tests/e2e/** must cover all required US; tests/api/\*\* all required CON-API. Each TC is covered from the directory its declared Level routes to (L3/Integration -> tests/integration/**, L4/API -> tests/api/\*\*, L5/E2E -> tests/e2e/\*\*; no declared Level -> tests/integration/\*\*).
+- Forbidden references guard the test-layer policy: a TC annotation outside its declared home is rejected — tests/api/** must not carry QFAI:SPEC-XXXX:TC-YYYY unless that TC declares L4/API, and tests/e2e/** likewise unless it declares L5/E2E.
 - Floor / ratio signals are planning hints, never gates; legacy scenario.feature / coverage ledger files remain optional inputs.

package/assets/init/.qfai/assistant/skills/qfai-atdd/references/test-case-depth-checklist.md

@@ -56,6 +56,22 @@ When multiple conditions interact:
 - [ ] Every BR-\* referenced in 04_Business-Rules.md has at least one positive and one negative test case.
 - [ ] Conditional business rules have test cases for each branch.
 
+## 8. Oracle Strength (オラクル強度)
+
+The seven sections above count **case categories**. This one asks whether each
+case's assertion can fail. A test that cannot fail satisfies every category.
+
+- [ ] No assertion uses truthiness where the boundary's own value is available
+      (`toBeTruthy()` on a result whose expected value is in hand).
+- [ ] No assertion compares a stored value against the same helper the fixture
+      wrote it with — that holds for every implementation, including a wrong one.
+- [ ] No loop asserts over a collection that is empty by construction.
+- [ ] No assertion observes a field the transport discards or a log the harness
+      swallows.
+- [ ] No assertion merely verifies a mock was called with what the test passed it.
+- [ ] Each case has a named production mutation that makes it fail, or a recorded
+      `equivalent-mutant` with the weaker contract clause named.
+
 ---
 
 ## Coverage Depth Matrix (テンプレート)
@@ -63,14 +79,17 @@ When multiple conditions interact:
 Reviewers and test-design-analysts MUST produce this matrix for each spec under review.
 Mark each cell: ✅ covered, ⚠️ partial, ❌ missing.
 
-| US/TC ID | Normal path | Error path | Boundary values | Special values | State transitions | Combinatorial | Status |
-| -------- | ----------- | ---------- | --------------- | -------------- | ----------------- | ------------- | ------ |
-| US-0001  | ✅/⚠️/❌    | ✅/⚠️/❌   | ✅/⚠️/❌        | ✅/⚠️/❌       | ✅/⚠️/❌          | ✅/⚠️/❌      | —      |
-| TC-0001  | ✅/⚠️/❌    | ✅/⚠️/❌   | ✅/⚠️/❌        | ✅/⚠️/❌       | ✅/⚠️/❌          | ✅/⚠️/❌      | —      |
+| US/TC ID | Normal path | Error path | Boundary values | Special values | State transitions | Combinatorial | Oracle strength | Status |
+| -------- | ----------- | ---------- | --------------- | -------------- | ----------------- | ------------- | --------------- | ------ |
+| US-0001  | ✅/⚠️/❌    | ✅/⚠️/❌   | ✅/⚠️/❌        | ✅/⚠️/❌       | ✅/⚠️/❌          | ✅/⚠️/❌      | ✅/⚠️/❌        | —      |
+| TC-0001  | ✅/⚠️/❌    | ✅/⚠️/❌   | ✅/⚠️/❌        | ✅/⚠️/❌       | ✅/⚠️/❌          | ✅/⚠️/❌      | ✅/⚠️/❌        | —      |
 
 ### Evaluation criteria
 
 - **PASS**: All cells are ✅ or ⚠️ with documented rationale for partial coverage.
+- **Oracle strength is not waivable by category coverage.** A row whose six
+  category cells are ✅ and whose Oracle strength cell is ❌ is a REVISE: it has
+  cases in every category and no evidence that any of them can fail.
 - **REVISE**: Any cell is ❌ without an explicit justification (e.g., Decision Record).
 
 ### Usage

package/assets/init/.qfai/assistant/skills/qfai-configure/SKILL.md

@@ -52,7 +52,7 @@ When unsure, read inputs in this order:
 
 - P1: `.qfai/assistant/constitution/*`
 - P2: `.qfai/assistant/manifest/*` + `.qfai/assistant/catalog/*`
-- P3: `.qfai/specs/<spec-id>/09_delta.md` (Decision Records; if no spec yet, state "not applicable")
+- P3: `.qfai/specs/<spec-id>/07_Decisions.md` + `.qfai/specs/_policies/08_Decisions.md` (Decision Records, `DR-*`; if no spec yet, state "not applicable")
 - P4: other artifacts (01_Spec.md, contracts, evidence, optional legacy `scenario.feature` / coverage ledgers)
 
 ## Sub-agent Delegation (MANDATORY)
@@ -70,7 +70,7 @@ Follow `.qfai/assistant/constitution/shared-skill-delegation-baseline.md`.
 ### Delegation Failure (Hard Stop)
 
 - No additional overrides.
-- Do not simulate roles. If the first required delegation fails, stop the stage and report remediation.
+- Do not simulate roles. Classify the failure per the baseline taxonomy first: `unavailable` stops the stage with a remediation report; `saturated` uses the bounded retry branch and keeps the stage open.
 
 ### Work Orders Summary (MANDATORY evidence)
 
@@ -88,7 +88,7 @@ Use the shared schema.
 - Follow `.qfai/assistant/constitution/shared-skill-delegation-baseline.md#reviewer-gate-baseline`.
 - Reviewer checks:
   - required roles were delegated;
-  - doctor evidence exists: `qfai doctor --fail-on error` completed without failing checks;
+  - doctor evidence exists: `npx qfai doctor --fail-on error` completed without failing checks;
   - Drift Protocol enforced;
   - test-layer policy enforced against `.qfai/assistant/catalog/test-layers.md`;
   - tool-count heuristics are signals, not gates.
@@ -106,7 +106,7 @@ Use the shared template.
 
 Use the shared template.
 
-- Required field: `Status (PASS/REVISE)`.
+- Required field: `Status (PASS/REVISE/PENDING)`. `PENDING` marks a gate that could not be run (see the baseline's reviewer-budget branch); it never counts as `PASS`.
 
 ## Stage 0 — Steering completion refresh (mandatory)
 
@@ -120,9 +120,9 @@ Follow `.qfai/assistant/constitution/shared-skill-operating-baseline.md#delta-re
 
 ## CRITICAL CONSTRAINTS (Read First)
 
-- Only update `qfai.config.yaml`, `.qfai/assistant/manifest/*` + `.qfai/assistant/catalog/*`, and `.qfai/evidence/configure-<run-id>.md` unless explicitly asked.
+- Only update `qfai.config.yaml`, `.qfai/assistant/manifest/*` + `.qfai/assistant/catalog/*`, `.qfai/assistant/constitution/quality.md` (gate capabilities only — never the surrounding rules), and `.qfai/evidence/configure-<run-id>.md` unless explicitly asked.
 - You MUST produce the required evidence file: `.qfai/evidence/configure-<run-id>.md`.
-  - `.qfai/evidence/` is intentionally NOT tracked by Git (it ships with a local `.gitignore`).
+  - `.qfai/evidence/` is intentionally NOT tracked by Git. The mechanism is the QFAI-managed block `npx qfai init` writes into the **project root** `.gitignore` (marker `# ── QFAI managed (generated by qfai init) ──`), which carries `.qfai/evidence/*`. A project generated by `npx qfai init` has no per-directory `.qfai/evidence/.gitignore`: nothing under the init asset set ships one, so an operator who needs to change what is tracked edits the root file, not a directory-local one. (qfai's own repository does carry that file, as a working-tree file outside the init assets — which is why the claim above used to read the other way.)
   - Do NOT commit evidence files; summarize key outcomes in the PR description instead.
 - You MUST run the mandatory checks listed below and record outcomes.
 - You MUST stop and escalate if tooling choices or runnable path remain ambiguous.
@@ -144,6 +144,10 @@ Note: /qfai-sdd includes a preflight step that bootstraps missing config/steerin
 
 - `qfai.config.yaml` is updated with a **minimal diff** focused on traceability globs.
 - `validation.traceability.testFileGlobs` reflects the real test layout.
+  - `npx qfai init` ships this empty on purpose. Detect the stack before setting it — `pyproject.toml` / `setup.cfg` (Python), `go.mod` (Go), `pom.xml` / `build.gradle` (JVM), `Cargo.toml` (Rust), `package.json` (JS/TS), `Gemfile` (Ruby), `composer.json` (PHP) — and derive globs from the test paths that actually exist, not from the language's convention alone.
+  - Cross-check against the test paths the spec ledgers declare (`.qfai/specs/*/tdd/test-list.md`, `06_Test-Cases.md`). Globs that do not cover the declared paths leave the traceability gate reporting success while covering nothing.
+  - The evidence file MUST show at least one matched file per declared test layer (unit / integration / api / e2e / component as applicable). A layer with zero matched files is a blocking gap, not a note.
+  - `npx qfai validate` emits `QFAI-TRACE-124` while the globs are unset or match zero files; the run is not done until it is gone.
 - `validation.traceability.testFileExcludeGlobs` is added only when needed.
 - If strict spec sections are explicitly requested, `validation.require.specSections` is updated with a minimal, evidence-based list.
 - A validation checklist with evidence (sample matched files) is produced.
@@ -304,7 +308,7 @@ Every 5 major actions, pause and restate:
 
 ## Constraints
 
-- Only update `qfai.config.yaml`, `.qfai/assistant/manifest/*` + `.qfai/assistant/catalog/*`, and `.qfai/evidence/configure-<run-id>.md` unless explicitly asked.
+- Only update `qfai.config.yaml`, `.qfai/assistant/manifest/*` + `.qfai/assistant/catalog/*`, `.qfai/assistant/constitution/quality.md` (gate capabilities only — never the surrounding rules), and `.qfai/evidence/configure-<run-id>.md` unless explicitly asked.
 - Do **not** modify tests or source code.
 - Avoid overly broad globs (e.g., `**/*`).
 - Exclude generated/output directories (`node_modules`, `.git`, `.qfai`, `dist`, `build`, `coverage`, `.next`, `out`, etc.).
@@ -321,6 +325,10 @@ Every 5 major actions, pause and restate:
 2. Read **project constitution / instructions** (if present):
    - `.qfai/assistant/constitution/constitution.md`
    - `.qfai/assistant/constitution/workflow.md` (or equivalent)
+   - `.qfai/assistant/constitution/quality.md` — the gate capability list. This
+     skill is its owner: reconcile it against the toolchain detected in step 3,
+     and record the discovered commands in `catalog/tech.md`. It had no reader
+     and no owner, so a stale statement there survived every configure pass.
 
 3. Inspect repo conventions:
    - package manager (pnpm/npm/yarn), test runner, lint/typecheck scripts, CI definitions

package/assets/init/.qfai/assistant/skills/qfai-discussion/SKILL.md

@@ -73,6 +73,10 @@ For UI-bearing targets, follow `references/design-dna-intake.md` while authoring
 
 Follow `.qfai/assistant/constitution/shared-skill-operating-baseline.md#gate-failure-autorepair-protocol` for validate, doctor, and quality-gate failures.
 
+The full completion logic, including the UI-bearing blocking conditions, is in
+`references/discussion-completion-matrix.md`. It must stay consistent with the canonical
+sidecar family declared above and with `templates/uiux/00_index.md#Forbidden Legacy Files`.
+
 Before declaring completion, you MUST:
 
 - verify all 15 mandatory output files exist and are populated;
@@ -80,7 +84,7 @@ Before declaring completion, you MUST:
 - ensure every deferred item has full metadata in `13_Deferred.md`;
 - ensure `02_Inception-Deck.md` and `03_Story-Workshop.md` include Mermaid diagrams;
 - ensure the UI-bearing sidecar family is complete and the root `DESIGN.md` draft exists at the consuming-project root and parses as valid front-matter;
-- run `qfai validate --profile discussion --fail-on error` and fix discussion-owned findings;
+- run `npx qfai validate --profile discussion --fail-on error` and fix discussion-owned findings;
 - avoid selecting a single visual winner in discussion artifacts.
 
 ### Reviewer Gate (MUST)
@@ -92,7 +96,7 @@ Reviewer checks must confirm:
 - discussion stayed planner-first and did not choose a single visual winner;
 - Drift Protocol is enforced; review policy is checked against `.qfai/assistant/catalog/test-layers.md`;
 - planning and coverage heuristics are signals, not gates;
-- review findings end with `Status (PASS/REVISE)` and Reviewer result is explicit as `PASS` or `REVISE`.
+- review findings end with `Status (PASS/REVISE/PENDING)` and Reviewer result is explicit as `PASS` or `REVISE` (`PENDING` marks a gate that could not be run and never counts as `PASS`).
 
 ## Sub-agent Delegation (MANDATORY)
 
@@ -109,11 +113,11 @@ Follow `.qfai/assistant/constitution/shared-skill-delegation-baseline.md`.
 ### Delegation Failure (Hard Stop)
 
 - No additional overrides.
-- Do not simulate roles. If the first required delegation fails, stop the stage and report remediation.
+- Do not simulate roles. Classify the failure per the baseline taxonomy first: `unavailable` stops the stage with a remediation report; `saturated` uses the bounded retry branch and keeps the stage open.
 
 ## Work Orders Summary
 
-Use the shared schema (per-row `Status (PASS/REVISE)` column, reviewer response `Result: PASS | REVISE`).
+Use the shared schema (per-row `Status (PASS/REVISE/PENDING)` column, reviewer response `Result: PASS | REVISE`).
 
 ## Completion Message & Next Actions (MUST)
 

package/assets/init/.qfai/assistant/skills/qfai-discussion/references/design-md-brand-catalog.md

@@ -83,8 +83,8 @@ Each archetype supplies defaults that the agent maps to the project's design sys
 
 Use this catalog during Step 11.3 Phase A:
 
-1. Score each archetype against the taste-interview results from `uiux/11_design_taste_interview.md`.
+1. Score each archetype against the brand intent captured in root `DESIGN.md` front-matter (`brand.voice`, `audience.emotion`, `audience.do_not_look_like`).
 2. Apply the archetype tie-breaker when two archetypes tie: highest visual-theme weight wins, then alphabetical name.
-3. Record the selected archetype in `uiux/12_design_system.md` under `## Visual Theme`.
+3. Record the selected archetype in root `DESIGN.md` front-matter as `brand.archetype`.
 4. The selected archetype's `aesthetic_properties` become the defaults for Color Palette, Typography, Spacing, and Animation sections.
 5. Step 11.3 Phase B then customizes those defaults to project-specific overrides.

package/assets/init/.qfai/assistant/skills/qfai-discussion/references/discussion-completion-matrix.md

@@ -6,17 +6,27 @@ Use this file for the full completion logic behind `/qfai-discussion`.
 
 Completion is blocked until all are true:
 
-1. Strategy selected
-2. Design taste interview completed
-3. 3-layer scoring family completed
-4. Dynamic overrides documented
-5. Option comparison completed
-6. Selected anchor documented
-7. Screen contracts drafted
+1. Root `DESIGN.md` exists at the consuming-project root and parses as valid front-matter
+   (`brand`, `audience`, and the full `visual.*` token tree).
+2. `# Brand Philosophy` body documents do/don't, brand signals, and exploration references
+   framed as **deviate-from** inputs.
+3. The canonical `uiux/` family is complete: `00_index.md`, `40_screen_contracts.md`,
+   `50_review_input_bundle.md`.
+4. Every screen contract in `40_screen_contracts.md` carries the full template schema.
+5. Exploration directions are carried unranked — no single visual winner is selected and the
+   design system is not finalized here (discussion is planner-first).
+6. No forbidden legacy sidecar exists under `uiux/` (see
+   `templates/uiux/00_index.md#Forbidden Legacy Files`).
+7. `Disposition: open` count is zero in `11_OQ-Register.md`.
+
+Evaluation axes are global constants (4-step ordinal: weak / acceptable / strong /
+exceptional) and are NOT authored as discussion sidecars, so no scoring, override, strategy,
+taste-interview, option-comparison or selected-anchor sidecar is required or permitted.
 
 ## Non-UI Packs
 
 - No uiux sidecars required
+- No root `DESIGN.md` required
 - No additional UI-bearing completion conditions apply
 - `prototyping.yaml` is not required
 

package/assets/init/.qfai/assistant/skills/qfai-discussion/references/rcp_footer.md

@@ -21,15 +21,21 @@
 - 条件付き reviewer:
   - `architecture-reviewer` — architecture-affecting decision がある場合
   - `product-surface-reviewer` — UI-bearing の場合
-- 各レビューは `PASS` / `FAIL` を返す
-- `FAIL` が出たら、失敗 reviewer と修正影響を受ける reviewer のみ再実行する
+- 各レビューは `PASS` / `REVISE` を返す（in-flight のレビュアー応答の語彙。SSOT は
+  `.qfai/assistant/constitution/shared-skill-delegation-baseline.md` の `## Reviewer response template`）
+- `REVISE` は review pack の `summary.json` を書き出す際に `status: "FAIL"` へ対応付けられる（同ファイルの
+  `### Verdict vocabulary` を参照）。第3の判定値を作らないこと
+- `REVISE` が出たら、失敗 reviewer と修正影響を受ける reviewer のみ再実行する
 
 ---
 
 ## Validate Hard Gate（必須）
 
-- 各 review cycle で `qfai validate --profile discussion --fail-on error --format github` を実行していること
-- `.qfai/report/validate.log` が存在し、最新の成果物に対応していること
+- 各 review cycle で `npx qfai validate --profile discussion --fail-on error --format github` を実行していること
+- `<paths.outDir>/validate.log`（既定 `.qfai/report/validate.log`）が存在し、最新の成果物に対応していること
+  - このファイルは `npx qfai validate` が実行のたびに自動で書き出す。`| tee` などのシェルリダイレクトは不要（PowerShell では動作しないため使用しない）
+  - `<paths.outDir>` は `qfai.config.yaml` の `paths.outDir`。既定以外に変更したプロジェクトでは `<paths.outDir>/validate.log` と `<paths.outDir>/run-*/` を見ること
+  - 最新性は `validate.log` の `run_log:` 行が最新の `run-*/` を指しているかで確認する
 
 ---
 

package/assets/init/.qfai/assistant/skills/qfai-discussion/references/review-cycle-playbook.md

@@ -13,7 +13,7 @@ Use this file for detailed review-pack handling in `/qfai-discussion`.
 1. Create a new review pack for each cycle.
 2. Apply the footer SSOT from `rcp_footer.md`.
 3. Run only the routed reviewers for the current phase and conditions.
-4. On `FAIL`, rerun only the failed reviewer and any reviewer whose scope changed.
+4. On `REVISE`, rerun only that reviewer and any reviewer whose scope changed. `REVISE` is the in-flight verdict (`.qfai/assistant/constitution/shared-skill-delegation-baseline.md#reviewer-response-template`); it becomes `status: "FAIL"` only when the pack's `summary.json` is written.
 5. Mark fixed only when all routed blocking reviewers are `PASS`.
 
 ## summary.json

package/assets/init/.qfai/assistant/skills/qfai-discussion/references/ui-bearing-playbook.md

@@ -39,7 +39,7 @@ the screen-level UX inputs:
 
 Trend Scan is **not** a sidecar file. Trend Scan lives at
 `04_Sources.md#Trend Scan` (part of the core 15-file discussion pack).
-The legacy `uiux/20_trend_scan.md` template was removed; `uiux/20_design_eval_invariant.md`
-is a different axis family (design evaluation invariant criteria) and does
-not own Trend Scan content. UI-bearing completion checks read Trend Scan
-from `04_Sources.md`.
+The legacy `uiux/20_trend_scan.md` template and the whole `uiux/20-24`
+design-evaluation family were removed; evaluation axes are now global
+constants and are not authored as discussion sidecars. UI-bearing
+completion checks read Trend Scan from `04_Sources.md`.

package/assets/init/.qfai/assistant/skills/qfai-discussion/references/ui_ux/review_audit_playbook.md

@@ -97,7 +97,7 @@ A good discussion artifact should make these reviewable without guesswork:
 ## Recommended Audit Order
 
 1. Validate user goal and flow.
-2. Validate option comparison and anchor choice.
+2. Validate that root `DESIGN.md` is authored and that exploration directions stay unranked (no winner selected here).
 3. Validate state coverage and contracts.
 4. Validate accessibility and platform fit.
 5. Validate anti-pattern absence.

package/assets/init/.qfai/assistant/skills/qfai-discussion/references/ui_ux/trend_scan_playbook.md

@@ -43,7 +43,7 @@ Always translate a reference into:
 
 - whether it fits the product's users and task frequency;
 - whether it improves or harms clarity;
-- whether it changes scoring in `uiux/20-24`;
+- whether it changes the brand direction recorded in root `DESIGN.md`;
 - whether it affects the selected anchor or contracts;
 - whether it should be adopted, adapted, or rejected.
 

package/assets/init/.qfai/assistant/skills/qfai-discussion/references/ui_ux_best_practices.md

@@ -93,18 +93,28 @@ Move volatile or bulky detail into the `ui_ux/` sidecar references and research
 
 ## QFAI-Specific Authoring Guidance
 
-### Use the sidecar family as the design SSOT for UI-bearing work
+### Use root `DESIGN.md` plus the canonical sidecar family as the design SSOT for UI-bearing work
 
-For UI-bearing packs, core design direction should resolve through:
+Brand-level direction is no longer authored as sidecars. For UI-bearing
+packs it resolves through:
 
-- strategy: `uiux/10_implementation_strategy.md`
-- taste constraints: `uiux/11_design_taste_interview.md`
-- scoring axes: `uiux/20-24`
-- option decision: `uiux/30_option_comparison.md`
-- selected anchor: `uiux/31_selected_anchor_screen.md`
+- brand SSOT (product intent, brand signals, anti-goals, deviate-from reference pool):
+  `<consuming-project-root>/DESIGN.md` (front-matter tokens + `# Brand Philosophy` body)
+- sidecar manifest: `uiux/00_index.md`
 - screen behavior contracts: `uiux/40_screen_contracts.md`
 - review handoff: `uiux/50_review_input_bundle.md`
 
+Those three are the entire canonical sidecar family. Evaluation axes are global constants
+(4-step ordinal: weak / acceptable / strong / exceptional, in
+`core/prototyping/evaluatorReview.ts#ORDINAL_AXES`) and are not authored as discussion
+sidecars. Discussion is planner-first: it carries exploration directions unranked and does
+not select a single visual winner, so there is no strategy, taste-interview, scoring,
+override, option-comparison or selected-anchor sidecar. Do NOT create
+`10_implementation_strategy.md`, `11_design_taste_interview.md`, the `20-24`
+design-evaluation family, `30_option_comparison.md`, or `31_selected_anchor_screen.md` —
+those filenames are forbidden, see `templates/uiux/00_index.md#Forbidden Legacy Files`, and
+creating them fails validation.
+
 ### What must be explicit in the discussion pack
 
 - why this design direction exists

package/assets/init/.qfai/assistant/skills/qfai-discussion/templates/01_Context.md

@@ -1,6 +1,6 @@
 # 01 Context
 
-<!-- UX-INTENT: If UI-bearing, see uiux/10_implementation_strategy.md for implementation strategy context -->
+<!-- UX-INTENT: If UI-bearing, see root DESIGN.md for brand intent and product context -->
 
 ## UI-bearing Classification
 

package/assets/init/.qfai/assistant/skills/qfai-discussion/templates/02_Inception-Deck.md

@@ -43,7 +43,7 @@ flowchart LR
     App --> Data["Data Store"]
 ```
 
-<!-- UX-INTENT: If UI-bearing, reference uiux/30_option_comparison.md and uiux/31_selected_anchor_screen.md for design option alignment -->
+<!-- UX-INTENT: If UI-bearing, reference root DESIGN.md and uiux/40_screen_contracts.md for design direction alignment -->
 
 ## 7. What Keeps Us Up at Night (Risks)