@sun-asterisk/sungen 3.2.12-beta.5 → 3.2.12-beta.7

package/dist/orchestrator/templates/ai-src/commands/create-data-test.md

@@ -94,9 +94,17 @@ Run (local-first): `[ -x ./bin/sungen.js ] && ./bin/sungen.js data gen --screen
 
 Run (local-first): `[ -x ./bin/sungen.js ] && ./bin/sungen.js data validate --screen <name> || npx sungen data validate --screen <name>`. This is stronger than lint (coverage): it checks every generated value is **semantically** correct — each `valid`/`boundary` value actually satisfies the constraints, and each synthesized `invalid` value actually violates them (a value in the `valid` set that breaks `maxLength`, or a "below min" case that landed inside the range, is a `VALUE_NOT_VALID` error). Fix any `error`; `INVALID_LOOKS_VALID` warns are a prompt to confirm the value is invalid for a non-length reason (charset/format). Each generated value also carries an `expected: valid|invalid` field in the test-data for traceability.
 
+### 5.6 Domain / stateful preconditions (spec + viewpoint only — no DB/API needed)
+
+If the spec describes a **stateful precondition** — an entity with relationships or a business state the test needs *before* it runs (e.g. "a paid order of user A", "a valid unused reset token ≤24h", "the session-storage buffer holds the entered values") — author a **state-recipe** in `qa/data-factory/recipes.yaml` under `resources:` (each with `entity`, `provenance`, `dependsOn`, `desiredState`, `lifecycle`, `with:` values; plus `bindings:`). Provenance is **not only api/db** — pick `prior-flow` / `fixture` / `client-storage` / `server-context` / `manual` (use `api`/`db` only when a datasource exists). Then run `data state --recipe <name>`: it validates the graph and renders a standardized **manual precondition** block (`@manual:data-setup` when it can't be auto-provisioned) to weave into the scenario's `Background`. This is the domain/stateful improvement that works with **only spec + viewpoint** — the testcase states the required state precisely instead of a vague note; it auto-provisions later if an api/db datasource is added (no rewrite). For a **journey where the same data flows across screens** (setup → confirmation → complete), declare a `shared: { entity, screens: [...] }` so one dataset is used across all of them (the Cross-artifact Gate then asserts they match).
+
+### 5.7 Cross-artifact check (testcase ↔ data agree)
+
+Run (local-first): `… data crosscheck --screen <name>`. It verifies the testcase and its test-data agree: no dangling `{{var}}` (every referenced var has a value), any cross-screen **shared** dataset is identical across the journey, and a declared state-precondition is surfaced in the feature. Fix any `error`. This is the `data-only` harness profile — `data validate` (values correct) + `data crosscheck` (values match the test); together they are the deterministic Data Gate + Cross-artifact Gate.
+
 ### 6. Weave into scenarios (if create-test already ran)
 
-If the `.feature` exists, turn the generated invalid/boundary sets into data-driven `Scenario Outline` + `Examples` (`@cases`) referencing the values — following `sungen-gherkin-syntax`. Otherwise leave the standardized `test-data/<name>.yaml` for `/sungen:run-test` to consume.
+If the `.feature` exists, turn the generated invalid/boundary sets into data-driven `Scenario Outline` + `Examples` (`@cases`) referencing the values — following `sungen-gherkin-syntax`. Weave any `data state` precondition block into the `Background`. Otherwise leave the standardized `test-data/<name>.yaml` for `/sungen:run-test` to consume.
 
 ## After running
 

package/dist/orchestrator/templates/ai-src/commands/create-test.md

@@ -167,7 +167,7 @@ If the unit is **api-first** (`qa/api/<name>/` or `qa/api/flows/<name>/`), the d
 
    **5d. Sequential fallback.** Use the single-context incremental path (Step 2: tier-by-tier `Write`/`Edit` batches) when: only **one** shard applies, **Copilot / no sub-agents**, or a constrained setup. Same output, just no speedup. **For flows**: `[Screen:Element]` namespace refs, test-data namespaced by phase, `@flow` tag.
 
-5.3. **Standardize field test-data — Data Factory (when the unit has input fields).** Don't hand-invent per-field values — make them **standard**. For each input field, author the field-map `qa/data-factory/<name>.fields.yaml` (field → catalog `type` + the spec's real `constraints` + `errorMap` placeholder→real code), then run (Bash, local-first) `[ -x ./bin/sungen.js ] && ./bin/sungen.js data lint --screen <name> || npx sungen data lint --screen <name>` and `… data gen --screen <name>`, per the **`sungen-data-factory`** skill. `data gen` **MERGES** into `test-data.yaml` (replaces only its field keys, preserves your scenario/namespaced data), so it is safe to run here. Weave the standardized **boundary/invalid** sets into `@cases` (`Scenario Outline` + `Examples`) with `CHK-*` trace; keep the map's `errorMap`/`constraints` aligned to the spec. **This is the same standard as `/sungen:create-data-test`, applied inline — so you do NOT run that command separately afterwards.** A unit with **no input fields** (navigation / list / capture-compare) → **skip this step** (nothing to standardize).
+5.3. **Standardize field test-data — Data Factory (when the unit has input fields).** Don't hand-invent per-field values — make them **standard**. For each input field, author the field-map `qa/data-factory/<name>.fields.yaml` (field → catalog `type` + the spec's real `constraints` + `errorMap` placeholder→real code), then run (Bash, local-first) `[ -x ./bin/sungen.js ] && ./bin/sungen.js data lint --screen <name> || npx sungen data lint --screen <name>` and `… data gen --screen <name>`, per the **`sungen-data-factory`** skill. `data gen` **MERGES** into `test-data.yaml` (replaces only its field keys, preserves your scenario/namespaced data), so it is safe to run here. Weave the standardized **boundary/invalid** sets into `@cases` (`Scenario Outline` + `Examples`) with `CHK-*` trace; keep the map's `errorMap`/`constraints` aligned to the spec. Then run `… data validate --screen <name>` (values are semantically correct) and — once the `.feature` exists — `… data crosscheck --screen <name>` (testcase ↔ data agree: no dangling `{{var}}`, shared datasets identical); fix any `error`. If the spec has a **stateful precondition** (an entity/state the test needs first), author a state-recipe + run `… data state` and weave its `@manual:data-setup` `Background` block (see the `sungen-data-factory` skill — works with spec+viewpoint only, no DB/API). **This is the same standard as `/sungen:create-data-test`, applied inline — so you do NOT run that command separately afterwards.** A unit with **no input fields** (navigation / list / capture-compare) → **skip this step** (nothing to standardize).
 
 5.4. **Depth self-check (deterministic — run BEFORE the audit).** Run `sungen depth-lint --screen <name>` (Bash). It reuses the audit's businessDepth classifier and splits every shallow business-critical scenario into two actionable buckets — act on them now so the audit/repair loop doesn't burn rounds on depth:
    - **DEEPEN IN PLACE** — add a real value assertion to each (`User see all [X] contain {{v}}`, `User remember [X] as {{v}}` + `… with {{v}}`). The printed `template` is a **hint** keyed off the theme — apply judgment to the scenario's actual claim; do NOT paste a value assertion that doesn't fit (e.g. a carousel-visibility scenario should assert the product SET, not a price). If a flagged scenario is genuinely visibility/behavior (not data-correctness), that's an over-count — leave it and note it, never fake an assertion.
@@ -196,7 +196,7 @@ If the unit is **api-first** (`qa/api/<name>/` or `qa/api/flows/<name>/`), the d
 4. Follow the `sungen-tc-generation` skill for section identification, viewpoint generation, and output format. **For flows**, use the "Flow Test Generation" section in the skill. When requirements exist, use the "Requirements-Driven Generation" strategy. **For Tier 1**, apply the **Lightweight Guard** — verify required fields, validation rules, business rules, security checks, and key state transitions all have TCs after generation. **For Tier 2+**, **MUST** apply the full **Mapping Contract** — walk every `spec.md` section top-to-bottom and produce the indicated TCs per Table 1; handle `test-viewpoint.md` per Table 2. Do not silently skip sections. Present sections as a numbered list and let user pick.
 5. Generate or update `.feature` + `test-data.yaml` following `sungen-gherkin-syntax` and `sungen-tc-generation` skills. Generate **group-by-group** (one viewpoint group at a time, tier-by-tier `Write`/`Edit` batches) to stay under the output-token cap. **For flows**: use `[Screen:Element]` namespace format, namespace test-data by phase, add `@flow` tag. **If a scenario needs `@query` DB verification**, check the datasource `engine` first — unsupported engine → follow `sungen-gherkin-syntax` § "Unsupported DB engine — fallback" instead of authoring `@query` steps.
    > **No parallel fan-out here.** Copilot has no sub-agents, so generation is sequential (the Claude Code variant fans out one `sungen-generator` per viewpoint group and merges). Same output, no speedup.
-5.3. **Standardize field test-data — Data Factory (when the unit has input fields).** For each input field, author the field-map `qa/data-factory/<name>.fields.yaml` (field → catalog `type` + real `constraints` + `errorMap`), then run `sungen data lint --screen ${input:name}` and `sungen data gen --screen ${input:name}` per the **`sungen-data-factory`** skill. `data gen` **MERGES** into `test-data.yaml` (replaces only its field keys, preserves the rest), so it is safe here. Weave the standardized boundary/invalid sets into `@cases` with `CHK-*` trace. **Same standard as `/sungen:create-data-test`, applied inline — do NOT run that command separately afterwards.** No input fields (navigation/list) → skip.
+5.3. **Standardize field test-data — Data Factory (when the unit has input fields).** For each input field, author the field-map `qa/data-factory/<name>.fields.yaml` (field → catalog `type` + real `constraints` + `errorMap`), then run `sungen data lint --screen ${input:name}` and `sungen data gen --screen ${input:name}` per the **`sungen-data-factory`** skill. `data gen` **MERGES** into `test-data.yaml` (replaces only its field keys, preserves the rest), so it is safe here. Weave the standardized boundary/invalid sets into `@cases` with `CHK-*` trace. Then `… data validate` (values correct) + `… data crosscheck` (testcase↔data agree); for a stateful precondition, author a state-recipe + `… data state` and weave its `@manual:data-setup` `Background` (works spec+viewpoint only). **Same standard as `/sungen:create-data-test`, applied inline — do NOT run that command separately afterwards.** No input fields (navigation/list) → skip.
 5.4. **Depth self-check (deterministic — BEFORE the audit).** Run `sungen depth-lint --screen ${input:name}`. It splits every shallow business-critical scenario into **DEEPEN IN PLACE** (add a real value assertion — the printed `template` is a theme-keyed hint, apply judgment to the actual claim; never fake one onto a visibility/behavior scenario) and **CROSS-SCREEN** (route to a flow / tag `@manual:Mx` + reason — removes it from the depth denominator honestly). Act on both, re-run until `deepen` is empty (or only honest over-counts remain), THEN gate. Lifts first-pass `businessDepth` mechanically instead of via 2–3 repair rounds.
 5.5. **Quality gate & repair (harness — always run).** Per `sungen-harness-audit`: run `sungen audit --screen ${input:name}` (structural), THEN do an **independent semantic review inline** using the `sungen-reviewer` criteria (does each scenario's steps PROVE its title/viewpoint? observable Thens? business-critical assertion depth?). Merge both sets of issues; if gate FAILs / findings exist, repair (budget 3) and re-audit — GATE missing theme → generate it (cross-screen → **automate it in the flow** via `/sungen:add-flow`, NOT a full `@manual` screen duplicate — `sungen audit` flags an automatable `@manual` as `MANUAL-AUTOMATABLE`; reserve `@manual:Mx` for true judgment/missing-capability); DEPTH → add data assertions; BALANCE → add business-core first; TRACE → align VP ids. Never fake a pass.
 5.6. **Record.** `sungen manifest --screen ${input:name}`. Ledger **each phase** (not just repair) — pick one `runId` at the start and pass it so `trace`/`ledger report` show THIS run, not a mix: `sungen ledger record --screen ${input:name} --run <runId> --step <discovery|viewpoint|gherkin|audit|repair:N> --ms <elapsed>`. On re-run, start with `sungen manifest --screen ${input:name} --diff` and only regenerate changed sections.

package/dist/orchestrator/templates/ai-src/skills/sungen-data-factory/SKILL.md

@@ -100,8 +100,11 @@ field-map for per-screen tweaks; use `.overwrite` only for rules that apply proj
 For CSV/Excel import or bulk-upload testing, `sungen data fixture --screen <name> [--rows N] [--invalid]`
 turns the field-map into a fixture FILE under `qa/fixtures/`: `<name>.csv` (N valid rows; nested groups
 → dotted headers; `unique` fields row-indexed) that should import cleanly, and — with `--invalid` —
-`<name>.invalid.csv` (one seeded-bad row per field + a `_violation` column) for row-level rejection
-tests. Reference it from an `@api` upload (`files:`) or a UI file input.
+`<name>.invalid.csv` (one seeded-bad row per field, **same columns as the valid file so a strict
+importer fails on the data, not an extra column**) plus a **sidecar** `<name>.invalid.expected.yaml`
+that says which 0-based row must be rejected and why (`field`, `chk`, `reason`, `expected_error`).
+Reference the CSV from an `@api` upload (`files:`) or a UI file input, and assert row rejection against
+the sidecar.
 
 ## Verify before done — the phuong-phap §6 checklist
 
@@ -128,6 +131,32 @@ output + a `NOT_APPLICABLE` lint info. A `--profile` (`regression` default · `f
 `security-min`) and `--locale`/`--channel`/`--sink` on `data gen|validate|lint` select a reduced set;
 selection is deterministic. Prefer setting `required` in the field-map over disabling the empty case.
 
+## Domain / stateful preconditions (spec + viewpoint only — no DB/API)
+
+Field values are not enough when a test needs a **business state first** — "a paid order of user A",
+"a valid unused reset token ≤24h", "the session-storage buffer holds the entered values". Declare a
+**state-recipe** in `qa/data-factory/recipes.yaml` under `resources:` — each resource has an `entity`,
+a **`provenance`** (`prior-flow` | `fixture` | `client-storage` | `server-context` | `manual` |
+`api` | `db`), optional `dependsOn`, `desiredState`, `lifecycle` (`reusable`/`consumable`/`expiring`/
+`leased`/`mutable-state`/`unique-per-run`), and `with:` values — plus `bindings:` (`orderId: ${order.id}`).
+Run **`sungen data state`**: it validates the graph (cycles, dangling deps, bindings) and renders a
+standardized **manual precondition** block — tagged **`@manual:data-setup`** when it can't be
+auto-provisioned — to weave into the scenario's `Background`. **This works with only spec + viewpoint:
+provenance is usually `prior-flow`/`fixture`/`manual`, NOT DB/API.** If a datasource is later added and
+a resource uses `api`/`db` provenance (+ a `provider:` endpoint), the *same* recipe auto-provisions — no
+rewrite. For a **journey where the same data flows across screens** (setup → confirmation → complete),
+add `shared: { entity, screens: [...] }` so one dataset is used across all of them.
+
+## Cross-artifact check — testcase ↔ data agree
+
+`data validate` checks the data alone; **`sungen data crosscheck`** checks the testcase and its data
+AGREE: no dangling `{{var}}` (every referenced var has a value), a cross-screen `shared` dataset is
+identical across the journey (confirmation shows what setup entered), and a declared state-precondition
+is surfaced in the feature. Run it after the `.feature` exists; fix any `error`. Together
+`data validate` + `data crosscheck` are the **Data Gate + Cross-artifact Gate** (the `data-only`
+harness profile). Every `data gen` also stamps a `_fingerprint` (catalog+generator+context hash) so a
+reader/cache can tell whether the inputs that produced the data changed.
+
 ## Weaving into scenarios
 
 When a `.feature` exists, express invalid/boundary sets as data-driven `Scenario Outline` + `Examples`

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@sun-asterisk/sungen",
-  "version": "3.2.12-beta.5",
+  "version": "3.2.12-beta.7",
   "description": "Deterministic E2E Test Compiler - Gherkin + Selectors → Playwright tests",
   "main": "src/index.ts",
   "types": "src/index.ts",
@@ -39,8 +39,8 @@
     "@babel/types": "^7.28.5",
     "@cucumber/gherkin": "^37.0.0",
     "@cucumber/messages": "^31.0.0",
-    "@sungen/driver-data-factory": "3.2.12-beta.5",
-    "@sungen/driver-ui": "3.2.12-beta.5",
+    "@sungen/driver-data-factory": "3.2.12-beta.7",
+    "@sungen/driver-ui": "3.2.12-beta.7",
     "chalk": "^5.6.2",
     "commander": "^14.0.2",
     "dotenv": "^17.2.3",

package/src/orchestrator/templates/ai-src/commands/create-data-test.md

@@ -94,9 +94,17 @@ Run (local-first): `[ -x ./bin/sungen.js ] && ./bin/sungen.js data gen --screen
 
 Run (local-first): `[ -x ./bin/sungen.js ] && ./bin/sungen.js data validate --screen <name> || npx sungen data validate --screen <name>`. This is stronger than lint (coverage): it checks every generated value is **semantically** correct — each `valid`/`boundary` value actually satisfies the constraints, and each synthesized `invalid` value actually violates them (a value in the `valid` set that breaks `maxLength`, or a "below min" case that landed inside the range, is a `VALUE_NOT_VALID` error). Fix any `error`; `INVALID_LOOKS_VALID` warns are a prompt to confirm the value is invalid for a non-length reason (charset/format). Each generated value also carries an `expected: valid|invalid` field in the test-data for traceability.
 
+### 5.6 Domain / stateful preconditions (spec + viewpoint only — no DB/API needed)
+
+If the spec describes a **stateful precondition** — an entity with relationships or a business state the test needs *before* it runs (e.g. "a paid order of user A", "a valid unused reset token ≤24h", "the session-storage buffer holds the entered values") — author a **state-recipe** in `qa/data-factory/recipes.yaml` under `resources:` (each with `entity`, `provenance`, `dependsOn`, `desiredState`, `lifecycle`, `with:` values; plus `bindings:`). Provenance is **not only api/db** — pick `prior-flow` / `fixture` / `client-storage` / `server-context` / `manual` (use `api`/`db` only when a datasource exists). Then run `data state --recipe <name>`: it validates the graph and renders a standardized **manual precondition** block (`@manual:data-setup` when it can't be auto-provisioned) to weave into the scenario's `Background`. This is the domain/stateful improvement that works with **only spec + viewpoint** — the testcase states the required state precisely instead of a vague note; it auto-provisions later if an api/db datasource is added (no rewrite). For a **journey where the same data flows across screens** (setup → confirmation → complete), declare a `shared: { entity, screens: [...] }` so one dataset is used across all of them (the Cross-artifact Gate then asserts they match).
+
+### 5.7 Cross-artifact check (testcase ↔ data agree)
+
+Run (local-first): `… data crosscheck --screen <name>`. It verifies the testcase and its test-data agree: no dangling `{{var}}` (every referenced var has a value), any cross-screen **shared** dataset is identical across the journey, and a declared state-precondition is surfaced in the feature. Fix any `error`. This is the `data-only` harness profile — `data validate` (values correct) + `data crosscheck` (values match the test); together they are the deterministic Data Gate + Cross-artifact Gate.
+
 ### 6. Weave into scenarios (if create-test already ran)
 
-If the `.feature` exists, turn the generated invalid/boundary sets into data-driven `Scenario Outline` + `Examples` (`@cases`) referencing the values — following `sungen-gherkin-syntax`. Otherwise leave the standardized `test-data/<name>.yaml` for `/sungen:run-test` to consume.
+If the `.feature` exists, turn the generated invalid/boundary sets into data-driven `Scenario Outline` + `Examples` (`@cases`) referencing the values — following `sungen-gherkin-syntax`. Weave any `data state` precondition block into the `Background`. Otherwise leave the standardized `test-data/<name>.yaml` for `/sungen:run-test` to consume.
 
 ## After running
 

package/src/orchestrator/templates/ai-src/commands/create-test.md

@@ -167,7 +167,7 @@ If the unit is **api-first** (`qa/api/<name>/` or `qa/api/flows/<name>/`), the d
 
    **5d. Sequential fallback.** Use the single-context incremental path (Step 2: tier-by-tier `Write`/`Edit` batches) when: only **one** shard applies, **Copilot / no sub-agents**, or a constrained setup. Same output, just no speedup. **For flows**: `[Screen:Element]` namespace refs, test-data namespaced by phase, `@flow` tag.
 
-5.3. **Standardize field test-data — Data Factory (when the unit has input fields).** Don't hand-invent per-field values — make them **standard**. For each input field, author the field-map `qa/data-factory/<name>.fields.yaml` (field → catalog `type` + the spec's real `constraints` + `errorMap` placeholder→real code), then run (Bash, local-first) `[ -x ./bin/sungen.js ] && ./bin/sungen.js data lint --screen <name> || npx sungen data lint --screen <name>` and `… data gen --screen <name>`, per the **`sungen-data-factory`** skill. `data gen` **MERGES** into `test-data.yaml` (replaces only its field keys, preserves your scenario/namespaced data), so it is safe to run here. Weave the standardized **boundary/invalid** sets into `@cases` (`Scenario Outline` + `Examples`) with `CHK-*` trace; keep the map's `errorMap`/`constraints` aligned to the spec. **This is the same standard as `/sungen:create-data-test`, applied inline — so you do NOT run that command separately afterwards.** A unit with **no input fields** (navigation / list / capture-compare) → **skip this step** (nothing to standardize).
+5.3. **Standardize field test-data — Data Factory (when the unit has input fields).** Don't hand-invent per-field values — make them **standard**. For each input field, author the field-map `qa/data-factory/<name>.fields.yaml` (field → catalog `type` + the spec's real `constraints` + `errorMap` placeholder→real code), then run (Bash, local-first) `[ -x ./bin/sungen.js ] && ./bin/sungen.js data lint --screen <name> || npx sungen data lint --screen <name>` and `… data gen --screen <name>`, per the **`sungen-data-factory`** skill. `data gen` **MERGES** into `test-data.yaml` (replaces only its field keys, preserves your scenario/namespaced data), so it is safe to run here. Weave the standardized **boundary/invalid** sets into `@cases` (`Scenario Outline` + `Examples`) with `CHK-*` trace; keep the map's `errorMap`/`constraints` aligned to the spec. Then run `… data validate --screen <name>` (values are semantically correct) and — once the `.feature` exists — `… data crosscheck --screen <name>` (testcase ↔ data agree: no dangling `{{var}}`, shared datasets identical); fix any `error`. If the spec has a **stateful precondition** (an entity/state the test needs first), author a state-recipe + run `… data state` and weave its `@manual:data-setup` `Background` block (see the `sungen-data-factory` skill — works with spec+viewpoint only, no DB/API). **This is the same standard as `/sungen:create-data-test`, applied inline — so you do NOT run that command separately afterwards.** A unit with **no input fields** (navigation / list / capture-compare) → **skip this step** (nothing to standardize).
 
 5.4. **Depth self-check (deterministic — run BEFORE the audit).** Run `sungen depth-lint --screen <name>` (Bash). It reuses the audit's businessDepth classifier and splits every shallow business-critical scenario into two actionable buckets — act on them now so the audit/repair loop doesn't burn rounds on depth:
    - **DEEPEN IN PLACE** — add a real value assertion to each (`User see all [X] contain {{v}}`, `User remember [X] as {{v}}` + `… with {{v}}`). The printed `template` is a **hint** keyed off the theme — apply judgment to the scenario's actual claim; do NOT paste a value assertion that doesn't fit (e.g. a carousel-visibility scenario should assert the product SET, not a price). If a flagged scenario is genuinely visibility/behavior (not data-correctness), that's an over-count — leave it and note it, never fake an assertion.
@@ -196,7 +196,7 @@ If the unit is **api-first** (`qa/api/<name>/` or `qa/api/flows/<name>/`), the d
 4. Follow the `sungen-tc-generation` skill for section identification, viewpoint generation, and output format. **For flows**, use the "Flow Test Generation" section in the skill. When requirements exist, use the "Requirements-Driven Generation" strategy. **For Tier 1**, apply the **Lightweight Guard** — verify required fields, validation rules, business rules, security checks, and key state transitions all have TCs after generation. **For Tier 2+**, **MUST** apply the full **Mapping Contract** — walk every `spec.md` section top-to-bottom and produce the indicated TCs per Table 1; handle `test-viewpoint.md` per Table 2. Do not silently skip sections. Present sections as a numbered list and let user pick.
 5. Generate or update `.feature` + `test-data.yaml` following `sungen-gherkin-syntax` and `sungen-tc-generation` skills. Generate **group-by-group** (one viewpoint group at a time, tier-by-tier `Write`/`Edit` batches) to stay under the output-token cap. **For flows**: use `[Screen:Element]` namespace format, namespace test-data by phase, add `@flow` tag. **If a scenario needs `@query` DB verification**, check the datasource `engine` first — unsupported engine → follow `sungen-gherkin-syntax` § "Unsupported DB engine — fallback" instead of authoring `@query` steps.
    > **No parallel fan-out here.** Copilot has no sub-agents, so generation is sequential (the Claude Code variant fans out one `sungen-generator` per viewpoint group and merges). Same output, no speedup.
-5.3. **Standardize field test-data — Data Factory (when the unit has input fields).** For each input field, author the field-map `qa/data-factory/<name>.fields.yaml` (field → catalog `type` + real `constraints` + `errorMap`), then run `sungen data lint --screen ${input:name}` and `sungen data gen --screen ${input:name}` per the **`sungen-data-factory`** skill. `data gen` **MERGES** into `test-data.yaml` (replaces only its field keys, preserves the rest), so it is safe here. Weave the standardized boundary/invalid sets into `@cases` with `CHK-*` trace. **Same standard as `/sungen:create-data-test`, applied inline — do NOT run that command separately afterwards.** No input fields (navigation/list) → skip.
+5.3. **Standardize field test-data — Data Factory (when the unit has input fields).** For each input field, author the field-map `qa/data-factory/<name>.fields.yaml` (field → catalog `type` + real `constraints` + `errorMap`), then run `sungen data lint --screen ${input:name}` and `sungen data gen --screen ${input:name}` per the **`sungen-data-factory`** skill. `data gen` **MERGES** into `test-data.yaml` (replaces only its field keys, preserves the rest), so it is safe here. Weave the standardized boundary/invalid sets into `@cases` with `CHK-*` trace. Then `… data validate` (values correct) + `… data crosscheck` (testcase↔data agree); for a stateful precondition, author a state-recipe + `… data state` and weave its `@manual:data-setup` `Background` (works spec+viewpoint only). **Same standard as `/sungen:create-data-test`, applied inline — do NOT run that command separately afterwards.** No input fields (navigation/list) → skip.
 5.4. **Depth self-check (deterministic — BEFORE the audit).** Run `sungen depth-lint --screen ${input:name}`. It splits every shallow business-critical scenario into **DEEPEN IN PLACE** (add a real value assertion — the printed `template` is a theme-keyed hint, apply judgment to the actual claim; never fake one onto a visibility/behavior scenario) and **CROSS-SCREEN** (route to a flow / tag `@manual:Mx` + reason — removes it from the depth denominator honestly). Act on both, re-run until `deepen` is empty (or only honest over-counts remain), THEN gate. Lifts first-pass `businessDepth` mechanically instead of via 2–3 repair rounds.
 5.5. **Quality gate & repair (harness — always run).** Per `sungen-harness-audit`: run `sungen audit --screen ${input:name}` (structural), THEN do an **independent semantic review inline** using the `sungen-reviewer` criteria (does each scenario's steps PROVE its title/viewpoint? observable Thens? business-critical assertion depth?). Merge both sets of issues; if gate FAILs / findings exist, repair (budget 3) and re-audit — GATE missing theme → generate it (cross-screen → **automate it in the flow** via `/sungen:add-flow`, NOT a full `@manual` screen duplicate — `sungen audit` flags an automatable `@manual` as `MANUAL-AUTOMATABLE`; reserve `@manual:Mx` for true judgment/missing-capability); DEPTH → add data assertions; BALANCE → add business-core first; TRACE → align VP ids. Never fake a pass.
 5.6. **Record.** `sungen manifest --screen ${input:name}`. Ledger **each phase** (not just repair) — pick one `runId` at the start and pass it so `trace`/`ledger report` show THIS run, not a mix: `sungen ledger record --screen ${input:name} --run <runId> --step <discovery|viewpoint|gherkin|audit|repair:N> --ms <elapsed>`. On re-run, start with `sungen manifest --screen ${input:name} --diff` and only regenerate changed sections.

package/src/orchestrator/templates/ai-src/skills/sungen-data-factory/SKILL.md

@@ -100,8 +100,11 @@ field-map for per-screen tweaks; use `.overwrite` only for rules that apply proj
 For CSV/Excel import or bulk-upload testing, `sungen data fixture --screen <name> [--rows N] [--invalid]`
 turns the field-map into a fixture FILE under `qa/fixtures/`: `<name>.csv` (N valid rows; nested groups
 → dotted headers; `unique` fields row-indexed) that should import cleanly, and — with `--invalid` —
-`<name>.invalid.csv` (one seeded-bad row per field + a `_violation` column) for row-level rejection
-tests. Reference it from an `@api` upload (`files:`) or a UI file input.
+`<name>.invalid.csv` (one seeded-bad row per field, **same columns as the valid file so a strict
+importer fails on the data, not an extra column**) plus a **sidecar** `<name>.invalid.expected.yaml`
+that says which 0-based row must be rejected and why (`field`, `chk`, `reason`, `expected_error`).
+Reference the CSV from an `@api` upload (`files:`) or a UI file input, and assert row rejection against
+the sidecar.
 
 ## Verify before done — the phuong-phap §6 checklist
 
@@ -128,6 +131,32 @@ output + a `NOT_APPLICABLE` lint info. A `--profile` (`regression` default · `f
 `security-min`) and `--locale`/`--channel`/`--sink` on `data gen|validate|lint` select a reduced set;
 selection is deterministic. Prefer setting `required` in the field-map over disabling the empty case.
 
+## Domain / stateful preconditions (spec + viewpoint only — no DB/API)
+
+Field values are not enough when a test needs a **business state first** — "a paid order of user A",
+"a valid unused reset token ≤24h", "the session-storage buffer holds the entered values". Declare a
+**state-recipe** in `qa/data-factory/recipes.yaml` under `resources:` — each resource has an `entity`,
+a **`provenance`** (`prior-flow` | `fixture` | `client-storage` | `server-context` | `manual` |
+`api` | `db`), optional `dependsOn`, `desiredState`, `lifecycle` (`reusable`/`consumable`/`expiring`/
+`leased`/`mutable-state`/`unique-per-run`), and `with:` values — plus `bindings:` (`orderId: ${order.id}`).
+Run **`sungen data state`**: it validates the graph (cycles, dangling deps, bindings) and renders a
+standardized **manual precondition** block — tagged **`@manual:data-setup`** when it can't be
+auto-provisioned — to weave into the scenario's `Background`. **This works with only spec + viewpoint:
+provenance is usually `prior-flow`/`fixture`/`manual`, NOT DB/API.** If a datasource is later added and
+a resource uses `api`/`db` provenance (+ a `provider:` endpoint), the *same* recipe auto-provisions — no
+rewrite. For a **journey where the same data flows across screens** (setup → confirmation → complete),
+add `shared: { entity, screens: [...] }` so one dataset is used across all of them.
+
+## Cross-artifact check — testcase ↔ data agree
+
+`data validate` checks the data alone; **`sungen data crosscheck`** checks the testcase and its data
+AGREE: no dangling `{{var}}` (every referenced var has a value), a cross-screen `shared` dataset is
+identical across the journey (confirmation shows what setup entered), and a declared state-precondition
+is surfaced in the feature. Run it after the `.feature` exists; fix any `error`. Together
+`data validate` + `data crosscheck` are the **Data Gate + Cross-artifact Gate** (the `data-only`
+harness profile). Every `data gen` also stamps a `_fingerprint` (catalog+generator+context hash) so a
+reader/cache can tell whether the inputs that produced the data changed.
+
 ## Weaving into scenarios
 
 When a `.feature` exists, express invalid/boundary sets as data-driven `Scenario Outline` + `Examples`