@cratis/pi 7.33.0 → 7.34.1

package/package/corpus/skills/cratis-screenplay-toolchain/references/renderable-subset.md

@@ -2,7 +2,8 @@
 
 ## Contents
 
-- What Stage 4.24 admits
+- Current admission (Stage 4.73.1)
+- Historical: what Stage 4.24 admitted
 - Authoring consequences in renderable mode
 - Ownership, publication and recovery
 - Reading V5
@@ -30,12 +31,35 @@ render itself before claiming renderability: this table is a reading of the sour
 render result.
 
 Renderable is a strict subset of executable (`executable-subset.md`). Order of events in
-`cratis render`: the CLI binds the model with its **bundled Screenplay** (4.66.0 in cratis 3.28.2, so PLAY0268
-and friends appear first), then Stage admits or
-rejects members. An ESM v6 model binds but Stage refuses it whole with `STAGE-ESM-016`
-(probed on 3.28.2; an ESM v7 model with generated values or responses fails binding first on the bundled 4.66.0 with `PLAY0268`, and Stage 4.24.2 would refuse it with `STAGE-ESM-016` too, tracked in Stage#201; on before 3.28.2 it failed binding first, with PLAY0268).
+`cratis render`: the CLI binds the **whole** model with its **bundled Screenplay** (4.137.0 in cratis
+3.53.1, so PLAY0268 and friends appear first, even with `--scope`), then Stage admits or rejects
+members of the selected scope. (Historical, cratis 3.28.2: an ESM v6 model bound but Stage refused it
+whole with `STAGE-ESM-016`.)
 
-## What Stage 4.24 admits
+## Current admission (Stage 4.73.1)
+
+Read at Stage `v4.73.1` (`Documentation/guides/build-renderer-target.md`, `SemanticVersionFeatures.cs`,
+`SemanticCratisAdmission*.cs`) and cli `v3.53.1`; source reading, not a render run (the worked
+example in `cratis-screenplay-render-and-gap-fill` was probed). Stage admits ESM v1 to v10 by
+version; each construct it does not render refuses with its own code before any artifact is planned.
+
+| Area | Renders at 4.73.1 | Refused (code) |
+| --- | --- | --- |
+| Slice kinds | `StateChange`, `StateView` | `Automation`, `Translate`, every reaction, capture, application trigger and clock specification (`STAGE-ESM-001`/`-024`); a reaction that both produces and invokes (`-037`, Stage#207; Arc 22.72.0+ can run it, and the Arc the scaffold pins needs a proven override; hand-write one reactor, never split the reaction) |
+| Queries | optional keyed snapshot lookups; observable collections, keyed or unkeyed (since 4.49.0) | a snapshot list: `STAGE-ESM-010`; `filter`, `scoped to`, `performer`, composite `by`: `PLAY0268` at binding |
+| Authorization | role-only, authenticated, opaque pure policies, claim conditions with Uuid, Date, DateTime or text targets (numeric and Boolean claim matches render as always false; Stage 4.73.1 `IsSupportedClaimPath`), `not`/and/or; roles read through the identity's `RoleClaimType` | role-claim URIs, nested opaque policies, unresolved or non-portable policies, no single identifier, opaque-policy subjects that are not text, Uuid, Date or DateTime (`STAGE-ESM-015`; Stage 4.73.1 `IsTextualSubject`, a restriction separate from declarative claim targets) |
+| Event sources, streams, routes (ESM v8) | rendered, with routed specifications | reserved stored names, unformattable routed fixtures, unsupported routed replay (`STAGE-ESM-030`) |
+| Numbers, case tables | whole numbers as `long` from ESM v8; case tables | out-of-`int` literals below v8 (`STAGE-ESM-031`) |
+| Generated values and responses (v7) | not rendered | `STAGE-ESM-028`/`-029` (Stage#175) |
+| Specifications | as in Stage's `admission.md` | `then no readmodel` (`-027`); an unauthenticated caller carrying roles or claims, `$context.occurred`-derived fixed values asserted (`-011`) |
+| v9 and v10 | reducer observer `from` filters | public events and event-target projections (`-032`); foreign origins and event-source captures (`-033`); directed Translate (`-024`); reaction `runsAs` (`-034`); reaction `from` filters (`-035`); per-production route overrides (`-036`) |
+| Evolved events | initial revisions only | `STAGE-ESM-026` (Stage#204) |
+| Descriptions, documentation | XML `<summary>`/`<remarks>` doc comments only (Stage#178 closed) | characters invalid in XML: `CLI-RENDER-005`; never behavior |
+
+Still bind-refused (not Stage): `reads` and `concurrency` (`PLAY0271`), command `handler`s, `pii`/`secret`
+(`PLAY0268`). Other rows of the table below that name 4.24 behavior are history.
+
+## Historical: what Stage 4.24 admitted
 
 Models using `eventsource`, `stream` or command routes are not renderable: they fail
 binding (`PLAY0268`) and Stage refuses their ESM version (`STAGE-ESM-016`); rendered
@@ -80,11 +104,14 @@ legacy ids or drop protected ids into routing to fit a renderer (PLAY0515).
   readmodel`). The seeded when-less `then query` only checks a lookup over a given read-model
   state; it proves no projection (Stage `SemanticSpecificationAdmission.Callers.cs`).
 - Use `$context.occurred` only; no `$context.identity.*` mappings.
-- A list is a design-only query and blocks binding, and so rendering, of the whole application on its own. A keyed `XById` sibling does not change that: while any list query remains, nothing in the model binds, runs or renders. Keep a list the domain needs and report the capability gap.
+- A list query binds (unfiltered, or narrowed by one `by` key); Stage 4.73.1 renders it only as an
+  `observable` collection, and refuses a snapshot list with `STAGE-ESM-010`. `filter`, `scoped to` and
+  `performer` queries block binding of the whole application. Keep a list the domain needs and report
+  the capability gap; marking it `observable` or keying it is a modeling decision for the user.
 - A `handler` (with or without `implementation` or `hint`) never binds; the slice is
   gap-fill, never "customized into" rendering.
-- Descriptions and documentation are never rendered (Stage#178): a rule that exists only in
-  prose is unenforced in rendered code.
+- Descriptions and documentation render only as XML doc comments (Stage#178 closed), never as
+  behavior: a rule that exists only in prose is unenforced in rendered code.
 
 ## Ownership, publication and recovery
 
@@ -104,8 +131,9 @@ legacy ids or drop protected ids into routing to fit a renderer (PLAY0515).
   `Customizations/styles.css`. The planner never reads it, and it never binds a modeled
   reaction or capture. A customization never makes a rejected model renderable.
 - `cratis render` does **not** build, test or run the application.
-- Stage 4.51.3's scaffold pins .NET 10, Arc 22.50.5, Chronicle client/kernel 19.32.0
-  and emits a `.frontend/` React/Vite app (Components 4.14.0, Scene 4.10.0).
+- Stage 4.73.1's scaffold pins .NET 10 and Arc and Chronicle client/kernel releases
+  and emits a `.frontend/` React/Vite app (Components and Scene releases pinned too); these are not the latest
+  releases (`cratis-stage-rendering-and-sandbox` `references/rendered-application.md`).
   `CratisEmbeddedScreenplayEnabled=false`; `Directory.Packages.props` is managed,
   never edit it. `[ProtectedDecision]` / `DecisionRead<T>` are available to gap-fill,
   not Stage-emitted enforcement; projection/key admission and unit-of-work enrollment
@@ -117,7 +145,8 @@ V5 is four separate results (admission, publication, build, tests), plus runtime
 was exercised; formats and the Debug-only spec classes are in `verdicts.md`. Three cases
 stay distinct: admission failure (nothing rendered); an existing generated base plus
 separately authorized hand-written gap-fill; and a fully hand-written delivery. Never claim a
-whole-application V5 from a subset. Drift: compare successive `.cratis-render.json`
+whole-application V5 from a subset (`cratis render --scope` renders an admitted part; state the scope).
+Drift: `cratis render ... --check` (CLI 3.44.0+; exit 6 means pending changes), or compare successive `.cratis-render.json`
 `semanticRevision` values from renders with the same `--name` and inputs (see `versions.md`,
 "Revisions and descriptions"); not the MCP `modelRevision`, not `open-workspace.revision`.
 

package/package/corpus/skills/cratis-screenplay-toolchain/references/traps.md

@@ -6,8 +6,8 @@ scope; negative capability claims there are not current verdicts. Current routes
 bind as v8, v7 reaches Stage planning, reference tests have CLI/MCP routes, and
 exact numbers are documented in the grammar. Compiler syntax: `cheat-sheet.md`.
 
-1. Clean V1 is not executable: binding errors appear only through MCP (V2) or `render`. Neither `screenplay <folder>` nor `cratis screenplay validate` binds. [probed]
-2. The Screenplay samples validate clean but are not all executable; copying their idioms (list queries, `@pii`, `today`, `reads`, `$causedBy`, templates, performers) breaks V3. [probed]
+1. Clean V1 is not executable: binding errors appear through MCP (V2), `render` or `cratis screenplay validate --executable` (bundled compiler, writes nothing). Neither `screenplay <folder>` nor plain `cratis screenplay validate` binds. [probed]
+2. The Screenplay samples validate clean but are not all executable; copying their idioms (filtered queries, `@pii`, `today`, `reads`, `$causedBy`, templates, performers) breaks V3. [probed]
 3. Know the target mode (design/executable/renderable) before writing; state it.
 4. Plain `produces` without `for` means allocation in the ESM but "the identifier" in rendered code. Always write `for <identifier>`. [PLAY0478 probed]
 5. PLAY0029 is a warning: a misspelled keyword (or `type` inside a slice) drops the construct silently. Use `--warnings-as-errors`. [probed]
@@ -41,10 +41,10 @@ exact numbers are documented in the grammar. Compiler syntax: `cheat-sheet.md`.
 33. Which compiler matters: cratis before 3.28.2 bundle Screenplay 4.60.1, which rejects ESM v6 (Automation/Translate) and reports a false PLAY0285 on cascades; cratis 3.28.2 bundles 4.66.0 and behaves like the standalone tool. Check `cratis --version` and name the tool in every verdict. [probed on both tools]
 
 34. A `from <Event> key <property>` routes only if every such event carries that property: an event that frees or updates a row keyed by another stream's id must carry that id in its payload, or the row never changes. Whether the compiler reports a missing key property is unverified [to probe]; check by hand.
-35. `description`/`documentation` are authoring metadata: board, MCP and Prologue show them; `cratis render` never does (Stage#178). A rule only in prose is unenforced in rendered code; make it a rule layer, a spec or a recorded target requirement.
+35. `description`/`documentation` are authoring metadata: board, MCP and Prologue show them; `cratis render` emits them only as XML doc comments (Stage#178 closed; invalid XML characters are CLI-RENDER-005), never as behavior. A rule only in prose is unenforced in rendered code; make it a rule layer, a spec or a recorded target requirement.
 36. One unadmitted construct (operations, `numbers exact`, composite view keys, refusal/redelivery) blocks whole-model binding. Routing is admitted v8, unlike composite view keys (#599); current CLI admits through v10 to construct checks. Overrides replace the whole route, filters never match unrouted facts. Do not remove routing or authorization to render a model.
 37. Personas remain outside ESM bytes, but 4.127.0 specifications accept `given caller as <Persona>`: one authenticated deterministic policy witness, not every matching caller. Inspect the chosen `or` branch; use explicit callers when synthesis is refused.
-38. A rule that depends on stored state is `reads <View>` + `require <expr> message "..."` (stated intent; PLAY0268/0271 at binding are expected in design mode). The slice `description` must mark it NOT enforced in the model today (Screenplay#129/#209) and name the target: Arc `[ProtectedDecision]` + `DecisionRead<T>` (Arc v22.39.0 and later; available to gap-fill on Stage 4.51.3's Arc 22.50.5/Chronicle 19.32.0, not emitted protection), Chronicle DCB (`concurrency` scope), or a constraint where one fits. Forbidden: caller-supplied copies of state in `require`, boolean attestation inputs (`confirmsX == true`), rules hidden in `handler`/implementation-hint prose. An unguarded materialized read is unsafe for a protected decision.
+38. A rule that depends on stored state is `reads <View>` + `require <expr> message "..."` (stated intent; PLAY0268/0271 at binding are expected in design mode). The slice `description` must mark it NOT enforced in the model today (Screenplay#129) and name the target: Arc `[ProtectedDecision]` + `DecisionRead<T>` (Arc v22.39.0 and later; available to gap-fill on Stage 4.51.3's Arc 22.50.5/Chronicle 19.32.0, not emitted protection), Chronicle DCB (`concurrency` scope), or a constraint where one fits. Forbidden: caller-supplied copies of state in `require`, boolean attestation inputs (`confirmsX == true`), rules hidden in `handler`/implementation-hint prose. An unguarded materialized read is unsafe for a protected decision.
 
 Identity/compliance traps checked at v4.127.0 (`identity.md`, `concepts.md`, `events.md`,
 `purposes.md`, `diagnostics.md`):
@@ -65,7 +65,7 @@ Identity/compliance traps checked at v4.127.0 (`identity.md`, `concepts.md`, `ev
 
 The SKILL.md top 12: 1, 4, 5, 6, 7, 10, 12, 13, 17, 18, 20, 21 (+ mode/version notes there). Traps 39 to 44 are tool traps added at the 4.66.0 pin.
 39. A handler never binds: PLAY0268 "Command '<n>' handler requires a constrained implementation attachment", with a file, a fence, `implementation` or `hint` alike. `implementation` and `hint` are accepted on a command handler, an operation execute/compensate phase and (from 4.65.0) a command property named rule, and are rejected on concept rules, built-in property rules and whole-command `require`/`validate` bodies; on a handler they are authoring intent only (no execution, lock or confirmation). A handler slice is gap-fill, not a customization of a rendered slice. [source, probed]
-40. `numbers exact` parses on 4.66.0 to 4.68.0 but never binds (PLAY0268; "not admitted by any supported executable model (ESM) version yet") and is PLAY0001 on older compilers; it is absent from `grammar.md`. Never use it in a model or example meant to execute. [probed]
+40. `numbers exact` (the source mode) parses but never binds at any release through 4.139.0 (exact lowering of qualifying Double-mode literals is separate: v11) (PLAY0268; "not admitted by any supported executable model (ESM) version yet") and is PLAY0001 on older compilers; it is absent from `grammar.md`. Never use it in a model or example meant to execute. [probed]
 41. MCP roots bug (4.63.1 and earlier, including the 4.60.1 bundled by cratis before 3.28.2; not present in 4.66.0 or cratis 3.28.2, probed): a dynamic-root server answers the roots request with an `id: null` error right after `notifications/initialized`; hosts treat it as fatal. `open-workspace.path` does not help. Start with a fixed root (`screenplay mcp <folder>`, `cratis screenplay mcp` in a project with `.cratis/ai.json`) or use 4.63.2 or later (cratis 3.28.2 or later). [source, probed]
 42. A symlinked model path is refused by the MCP server (macOS `/tmp` is a symlink); pass the physical path. Proposals are connection-local (at most 16); `apply` persists identities to `.screenplay/identities.json` and is not crash-atomic across files; never retry `apply` after `ApplyOutcomeUnknown`. [source]
 43. `modelRevision` is canonical semantic (line positions and descriptions do not change it; a file move only if it changes logical placement or application identity) and, without `identities.json`, changes with the root folder name; it is not comparable with a render's `semanticRevision`. Source identity for verdicts is the commit plus the digest from the source-identity helper (`cratis-screenplay-modeling-lifecycle` `references/verdicts-and-modes.md` "Source identity"). [probed]

package/package/corpus/skills/cratis-screenplay-toolchain/references/verdicts.md

@@ -46,12 +46,17 @@ V3 are read from the MCP server or from a render, V1 and V5 from the CLIs.
 - Warnings are failures here. PLAY0029 (a construct dropped silently) and the unresolved-name
   warnings are the defects this verdict exists to catch.
 - Syntax and consistency checks (PLAY0282 to PLAY0294) run; binding does not.
-- Report: `V1 pass (screenplay 4.66.0, 3 files)`, `V1 fail: PLAY0029 a.play:12 ...`.
+- Report: `V1 pass (screenplay 4.139.0, 3 files)`, `V1 fail: PLAY0029 a.play:12 ...`.
 
 ## V2: executable diagnostics
 
 Needs an MCP connection (V2 and V3 belong to the session that owns MCP; a subagent without
-MCP reports `V2 not run: no MCP in this agent` unless the brief supplies fresh output).
+MCP reports `V2 not run: no MCP in this agent` unless the brief supplies fresh output). Without
+MCP, a binding check exists: `cratis screenplay validate --executable <root>` binds with the
+**bundled** compiler (CLI 3.53.1 bundles Screenplay 4.137.0, not the standalone 4.139.0), never
+renders or writes files, and exits 5 with "The model does not bind into an executable semantic
+model" when it does not. Report it as `V3 (bundled 4.137.0 via cratis validate --executable)`; it
+gives binding errors, not the V2 `executable-diagnostics` view, so V2 stays MCP-only.
 
 1. Start the server with a fixed root: `screenplay mcp <model-folder>` or `cratis screenplay
    mcp` in a project with `.cratis/ai.json` (see `versions.md`, roots bug). Send
@@ -62,7 +67,7 @@ MCP reports `V2 not run: no MCP in this agent` unless the brief supplies fresh o
    current `expectedRevision` (page with `offset`; paging beyond the first page needs the
    source revision).
 4. Report the codes with counts, blocking (error severity) apart from informational. A list
-   with blocking entries is a valid V2 result (`V2 read: 6 blocking, PLAY0268 x3 list query,
+   with blocking entries is a valid V2 result (`V2 read: 6 blocking, PLAY0268 x3 filter query,
    pii ...`); it is the evidence for a "V3 blocked" report. The view also lists
    information-severity diagnostics (PLAY0269 deferred UI, PLAY0270 authoring metadata such as
    descriptions, personas and screens) on models that bind; they do not block.
@@ -79,8 +84,8 @@ do not block. The ESM then binds (and `read-workspace view=executable-model` rep
 comparison needs it). Binding is what makes a model renderable at all. V3 means "binds",
 never "specifications pass".
 
-- Report the ESM version the model needs: `V3 ready (screenplay 4.68.0, ESM v6)`,
-  `V3 blocked: PLAY0268 x3 (list query, pii) - design scope kept`.
+- Report the ESM version the model needs: `V3 ready (screenplay 4.139.0, ESM v6)`,
+  `V3 blocked: PLAY0268 x3 (filter query, pii) - design scope kept`.
 - A blocked V3 is not repaired by deleting protection or domain rules. Never remove `pii`,
   `secret`, authorization, list queries, automations or rules to reach V3; record the
   slice and the code in the gap list.
@@ -137,11 +142,11 @@ What Stage admits: `renderable-subset.md`.
 One line per verdict, tool and version first, evidence second:
 
 ```text
-V1 pass (screenplay 4.66.0, 3 files)
-V2 read (screenplay 4.66.0 via MCP): 0 blocking, 3 informational (PLAY0270)
-V3 ready (screenplay 4.66.0, ESM v3)
+V1 pass (screenplay 4.139.0, 3 files)
+V2 read (screenplay 4.139.0 via MCP): 0 blocking, 3 informational (PLAY0270)
+V3 ready (screenplay 4.139.0, ESM v3)
 V4 not run: no route
-V5 admission pass (cratis 3.28.3, Stage 4.24.2); publication written 14 files; build not run; tests not run
+V5 admission pass (cratis 3.53.1, Stage 4.73.1); publication written 14 files; build not run; tests not run (a sample line, not a result)
 ```
 
 Rules: a clean verdict never implies a later one; a verdict names the **source identity** it

package/package/corpus/skills/cratis-screenplay-toolchain/references/versions.md

@@ -23,6 +23,8 @@ probe; older probes retain their original labels in the other references.
 | Stage bundled by CLI | **4.73.1** | Same file; Stage's own `Directory.Packages.props` also pins Screenplay 4.137.0 |
 | Cratis.Arc.Screenplay bundled by CLI | **22.54.0** | Same file; this is the generator package, not a blanket rendered-app dependency pin |
 | Chronicle / Fundamentals bundled by CLI | **19.32.0 / 7.22.8** | Same file |
+| Latest Arc / Chronicle / Fundamentals released when this was checked | **22.72.0 / 19.50.0 / 7.23.0** | Released tags; newer than the CLI bundle and than Stage's scaffold. Arc 22.72.0 and later run a reactor returning events and commands (absent through 22.71.0); the Stage 4.73.1 scaffold pins Arc 22.68.0, whose command handler accepts a collection only when every element is a command, so a rendered app needs a proven dependency override to 22.72.0 or later |
+| Stage 4.73.1 rendered backend and frontend | **Arc 22.68.0 / Chronicle 19.45.0**; Components 4.26.2, Scene 4.21.0, `@cratis/arc` 22.58.2, Fundamentals 7.22.8 | Stage `v4.73.1:Source/Rendering.Cratis/Scaffolding/CratisBackendApplicationScaffoldProfile.cs` and `CratisFrontendPackageSet.cs` (`Current`); not the latest releases and not bumped by hand |
 | Historical Stage 4.51.3 rendered backend | **Arc 22.50.5 / Chronicle 19.32.0** | Stage `v4.51.3:Source/Rendering.Cratis/Scaffolding/CratisBackendApplicationScaffoldProfile.cs`; client/testing/image match; scaffold disables embedded reverse extraction |
 
 The published `screenplay-contract.json` beside this reference is the machine-readable authority for PLAY codes, MCP tools and parameters, and standalone CLI commands/options. Its unchanged release bytes are pinned by `screenplay-contract.version` to the standalone version above. Before adding a claim, search this large JSON file by exact `code` in `diagnostics`, or by exact `name` in `mcpTools`/`cliCommands`, and inspect that object's retirement flag, parameters or options; do not load the full file into context. For example, search for `"code": "PLAY0029"` or `"name": "read-workspace"`; `screenplay-contract-exceptions.json` preserves explicitly reasoned historical statements, not current-release exceptions. Coverage gaps are mention inventories, not proof that a skill teaches a feature.
@@ -175,8 +177,9 @@ specifications, UI composition and `diagnostics.md`.
   Stage 4.24.1/4.24.2. Those old negative v7 probes do not describe CLI 3.41.0.
 - CLI before 3.28.2 bundled Screenplay 4.60.1: false cascade PLAY0285, v6 refusal
   and dynamic-root initialization bug. The roots bug was fixed in 4.63.2.
-- CLI 3.28.2 file validation ignored imports; this is an old probe, not a current
-  reason to invent missing declarations. Validate the complete root.
+- CLI 3.28.2 file validation ignored imports (cli#244, closed by PR #258): later CLI releases compile
+  the file as an application root together with its imports. Folder mode stays the recommended check;
+  validate the complete root.
 - Older standalone empty-folder probes exited 0. Current validation still needs
   a nonzero file count, and specification execution needs an expected selection.
 
@@ -189,7 +192,10 @@ preserve assigned identities; do not reconstruct them from names.
 
 A render manifest's `semanticRevision` is a different provenance value, made
 with the render name and bundled compiler. Do not compare it directly to MCP
-`modelRevision`. Compare successive renders with the same identity and inputs.
+`modelRevision`. Compare successive renders with the same identity and inputs, or use
+`cratis render ... --check` (CLI 3.44.0 or later). Descriptions and event documentation render as XML
+doc comments, so editing them changes managed-file hashes (a `--check` exit 6) but not the executable
+model revision.
 
 ## Upgrade checklist
 

package/package/corpus/skills/cratis-screenplay-toolchain/verification.json

@@ -25,7 +25,10 @@
     { "kind": "skill-contains", "value": "4.66.0" },
     { "kind": "skill-contains", "value": "references/generated-responses-example.md" },
     { "kind": "skill-contains", "value": "Operations and exact numeric" },
-    { "kind": "skill-contains", "value": "Stage generated-value/response refusals `STAGE-ESM-028`/`029`" },
+    { "kind": "skill-contains", "value": "`STAGE-ESM-028`/`029`, not a blanket compiler refusal" },
+    { "kind": "skill-contains", "value": "cratis screenplay validate --executable" },
+    { "kind": "skill-contains", "value": "XML doc comments" },
+    { "kind": "skill-contains", "value": "Stage 4.73.1 rendered backend", "file": "references/versions.md" },
     { "kind": "skill-contains", "value": "cratis screenplay validate" },
     { "kind": "skill-contains", "value": "cli#244" },
     { "kind": "skill-contains", "value": "`notifications/initialized`" },

package/package/corpus/skills/cratis-screenplay-ui-composition/LICENSE

@@ -5,5 +5,5 @@ Third-party notices
 
 This skill adapts material from agentic-engineer by Martin Dilger and Nebulit GmbH
 (https://github.com/Nebulit-GmbH/agentic-engineer, commit 07b0f30648d663cb588d7e2c7aa031af9dfc21f2;
-https://nebulit.de), used with the agreement of Martin Dilger and Nebulit GmbH. That repository
-carries no licence file; the adapted material remains credited to Martin Dilger and Nebulit GmbH.
+https://nebulit.de), used with the explicit consent of Martin Dilger and Nebulit GmbH. The adapted
+material remains credited to Martin Dilger and Nebulit GmbH.

package/package/corpus/skills/cratis-screenplay-ui-composition/SKILL.md

@@ -31,11 +31,9 @@ source is the single flow model.
 
 ## Verified product sources
 
-| Package | Version | Purpose |
-| --- | --- | --- |
-| `Cratis.Screenplay` | `4.31.0` | Original UI composition parser evidence |
-| `Cratis.Screenplay` | main `fd18129` | Cancellation routing and `optional`; changed example compiled |
-| `Cratis.Screenplay` | `4.66.0` (`c89198b`) | Statements in this skill re-verified: the nine `Documentation/screenplay/` pages above are unchanged since `v4.31.0` except `file-references.md` (typed-context and `implementation` paragraphs); `PLAY0269` in `Diagnostics/DiagnosticCodes.cs`; the new section compiled with `screenplay` 4.66.0 and `cratis screenplay validate` 3.28.2 |
+Current pins: toolchain `references/versions.md`; UI statements were re-read at Screenplay v4.139.0
+(`screens.md`, `contributions.md`, `forms.md`, `templates.md`, `ui-profile.md`, `fixtures/screen-release-ui.play`). Older provenance:
+`4.31.0` (original parser evidence), main `fd18129` (cancellation routing, `optional`), `4.66.0` (`c89198b`: `PLAY0269`).
 
 The update follows `commands.md`, `events.md`, `queries.md` and decision 0023
 at that main commit (after v4.52.0). It does not verify UI rendering.
@@ -222,9 +220,11 @@ module-level.
 `navigate to <Screen> by <param>` is **not** string interpolation; it reuses the
 same typed navigate binding a screen action uses.
 
-⚠️ The v4.31.0 baseline, re-verified at v4.66.0, ships `navigate`, `label` and `order` only. Grouping beyond a
-flat ordered list, and an explicit override for when nearest-enclosing is not the
-point you mean, are deliberately left for later. The screens-release vector
+⚠️ At v4.139.0 a navigation contribution also carries `id`, `icon`, `presentation`, `group` and
+`destination outlet <name> | dialog <DialogTemplate> | external "<route>"`; its `navigate to` takes
+`route`, `outlet` and `parameter` lines, and an undeclared outlet or dialog target is `PLAY0654`
+(`contributions.md`). An explicit override for when nearest-enclosing is not the point you mean is
+not in the language. The screens-release vector
 (canonical `ScreenComposition` corpus v1) adds typed screen composition; public
 CLI 3.43.0 renders it and passes the browser run. Report a refused form as a
 capability gap, never a flat-list fallback; Studio production proof is separate.
@@ -286,8 +286,8 @@ module Invoicing
 The inline event targets `invoiceId` without copying it into payload; the
 projection obtains it from event context. A command `returns` clause (scalar or
 record) binds and executes from standalone Screenplay 4.68.0 (ESM v7, `commands.md`,
-"Generated values and responses"); the 4.66.0 bundled in `cratis` 3.28.x reports `PLAY0268`, and
-Stage 4.24.2 does not render ESM v7 (`STAGE-ESM-016`, tracked in Stage#201). Form `on submit` and
+"Generated values and responses") and in the current cratis bundle (the 4.66.0 in 3.28.x reported `PLAY0268`);
+Stage 4.73.1 refuses it (`STAGE-ESM-029`; generated properties `-028`). Form `on submit` and
 interaction `on success` response-name scopes, failure clearing and response
 execution remain unavailable.
 An existing success continuation does not imply a response contract.

package/package/corpus/skills/cratis-screenplay-ui-composition/references/flow-review.md

@@ -36,7 +36,7 @@ the build.
 BerthListScreen
   Shows:      the berths booked so far (berth, boat name).
   Arrived by: the entry point, and after a booking is submitted.
-  Can do:     book a berth (BookBerth, opens BookBerthScreen).
+  Can do:     open the booking screen (a click on the booking section opens BookBerthScreen).
   Gaps:       none. Empty list: nothing booked yet.
 
 BookBerthScreen

package/package/corpus/skills/cratis-screenplay-ui-composition/references/provenance.md

@@ -6,11 +6,10 @@ No TrogonStack text is adapted in this skill. Nebulit material is credited in th
 |---|---|---|---|---|
 | Language constructs, rules and Verify list | Cratis/Screenplay v4.31.0 documentation, re-verified at v4.66.0 | MIT | original | SKILL.md |
 
-## Adapted closely (Martin Dilger and Nebulit GmbH, with agreement)
+## Adapted closely (Martin Dilger and Nebulit GmbH)
 
 Source: https://github.com/Nebulit-GmbH/agentic-engineer at commit `07b0f30648d663cb588d7e2c7aa031af9dfc21f2`,
-by Martin Dilger and Nebulit GmbH (https://nebulit.de). The repository carries no licence file; this material
-is adapted with the agreement of Martin Dilger and Nebulit GmbH.
+by Martin Dilger and Nebulit GmbH (https://nebulit.de), used with the explicit consent of Martin Dilger and Nebulit GmbH.
 
 | Source file | Used in | How |
 |---|---|---|

package/package/corpus/skills/cratis-screenplay-ui-composition/references/screen-composition-contract.md

@@ -17,7 +17,9 @@
 
 This reference is the authoring checklist for the screens-release UI surface. It
 records the contract an agent must look for in the installed Screenplay tool and
-in peer runtime/render contracts; it is not a second language definition. When a
+in peer runtime/render contracts; it is not a second language definition. The syntax
+below was re-read at Screenplay v4.139.0 (`screens.md`, `forms.md`, `templates.md`,
+`contributions.md`, `ui-profile.md` and `fixtures/screen-release-ui.play`); the shapes are released, not proposed. When a
 syntax form below is not present in `syntax-schema` or does not compile with the
 project's selected Screenplay version, report the support level instead of
 hand-writing an alternate shape.
@@ -55,22 +57,25 @@ Still true: a guarded Close action and a double-click interaction are refused
 (`STAGE-SCENE-ACTION-001`, `STAGE-SCENE-INTERACTION-001`, defined from Stage
 4.51 through 4.57) and recorded as unsupported, never emulated; Studio production
 Play proof is pending a test-account decision. Refusing an unresolvable UI
-profile and recovering a rendered application shipped in Cratis/cli#301. On Stage 4.24.2 or cratis 3.28.x
-renderers, authored screens are still reported as omitted/default composition.
+profile and recovering a rendered application shipped in Cratis/cli#301. (History: on Stage 4.24.2 or
+cratis 3.28.x renderers authored screens were reported as omitted/default composition; Stage 4.73.1
+(CLI 3.53.1) renders navigation contributions and toolbars, and `CLI-RENDER-015`, which refused them, is retired.)
 
 ## Data context and component bindings
 
 A component instance is a presentation of a typed Screenplay data context. The
-proposed syntax is:
+released syntax (canonical fixture `fixtures/screen-release-ui.play`) is:
 
 ```screenplay excerpt
 component <Package.Component> <instanceName>
-  context <binding>
-  property <property.path> from <binding>
+  id "<stable-id>"
+  context from query <Query>
+  property <property.path> from component <stable-id>.<outputPath> null preserve
+  property <property.path> from literal <value>
   property <property.path> = "literal"
   icon <IconName>
   presentation <key> "value"
-  exposes <name> from <binding>
+  exposes <name> from component <stable-id>.<outputPath>
   outlet <name>
     ...nested screen directives...
   on click
@@ -78,6 +83,11 @@ component <Package.Component> <instanceName>
       with <name> from <binding>
 ```
 
+A binding is `from query <Query>`, `from component <stableId>.<outputPath>` (quote the id when it has
+punctuation) or `from literal <value>`, and may carry `mode oneWay|twoWay`, `null propagate|clear|preserve`
+and `expected <Type>`. A component's `id` pins the stable instance id Studio and Stage use; without it the
+authored name is the id. Malformed bindings are diagnostics, and the raw text is kept.
+
 Rules for agents:
 
 - Discover the exact node shape through `syntax-schema` before proposing it; do
@@ -137,6 +147,15 @@ screen template MasterDetail
 - `outlet` declares an addressable child surface. It is different from a static
   slot such as `sidebar` or `main`.
 
+A template may also declare `content <slot>` (screen directives the template renders in that slot itself;
+the slot must be declared, `PLAY0629`; templates may not nest in a cycle, `PLAY0656`), `display` and
+`description` (what a picker shows), and `scopes application|module|feature|subfeature|slice|none`
+(use outside its scopes is `PLAY0631`). Which of a template's components a screen may configure is a
+top-level **exposure** (`exposure for <Owner>`, owner a layout or template, `PLAY0622`; `reexposes <Owner>`,
+`PLAY0623`/`PLAY0624`), and a screen or template configures it with a top-level `instance <Screen>`
+(`set <component>.<path> = <value>`, `items <component>.<path>` with `item <id>` lines; `PLAY0626` to
+`PLAY0628`). Layouts take `category`, `type`, `exposes` and `outlet` too.
+
 A module, feature or slice may select a default template for contained screens:
 
 ```screenplay
@@ -149,7 +168,10 @@ installed parser/printer supports the scoped `template` directive before using i
 
 ## Forms and columns
 
-A form remains bound to exactly one command. The screens-release additions make
+A form remains bound to exactly one command. It may carry one `description` line (report-only),
+`generation auto|manual` (Scene's form generation mode: generate fields from command metadata, or the
+authored fields are the contract), a `layout` block (column, gap and placement geometry, independent of
+`generation`) and the `columns` display-order hint. The screens-release additions make
 its generated fields explicit when the automatic shape is not enough:
 
 ```screenplay excerpt
@@ -182,7 +204,8 @@ ui profile Desktop
   packages
     Product.DesignSystem
     Cratis.Components
-  icons Product.Icons
+  icons
+    Product.Icons
 ```
 
 - Package order is still override-priority order. Later packages shadow earlier
@@ -204,13 +227,15 @@ toolbar InvoiceToolbar
     label "Register"
     icon Add
   item details navigate to InvoiceDetails
-    parameter invoiceId from selected.invoiceId
+    parameter invoiceId from component invoices.selectedItem.id
+  item compose dialog InvoiceComposer
 ```
 
-Use a toolbar for reusable groups of actions; use inline `action` when a single
-screen owns the button. Dialog opening stays an interaction (`on click open dialog
-<DialogTemplate>`) unless the installed language exposes a dedicated toolbar-dialog
-node. A dialog destination must state every input with a typed `with`/`parameter`
+A toolbar item is `item <name> action <Command>`, `item <name> navigate to <Screen>` or
+`item <name> dialog <DialogTemplate>`. Use a toolbar for reusable groups of actions; use inline `action`
+when a single screen owns the button. A parameter that reads a selection is a typed
+`from component <stableId>.<path>` binding (the bare `selected.invoiceId` form parses only as a legacy
+data-context path). A dialog destination must state every input with a typed `with`/`parameter`
 binding; a runtime default is not an authored contract.
 
 What Scene 4.17 adds to the runtime side of this contract, so an agent knows what a

package/package/corpus/skills/cratis-screenplay-ui-composition/references/worked-flow.md

@@ -1,7 +1,10 @@
 # Worked flow: marina berth bookings
 
-Complete document, design-mode (the list query `ListBerths` keeps it out of the executable
-subset; see `cratis-screenplay-slice-design`). Used by `SKILL.md`, *Compose in this order*.
+Complete document. It compiles and binds (screens, forms and descriptions are informational
+PLAY0269/0270; the `ListBerths` list query binds since Screenplay 4.105.0, and Stage 4.73.1 would render only
+an `observable` list). Used by `SKILL.md`, *Compose in this order*. Opening an input screen is a
+click (`section ... on click navigate to`); the input screen's `action` carries the destination
+**after success**.
 
 ```screenplay
 concept BerthId : Uuid
@@ -31,8 +34,10 @@ module Marina
       query ListBerths => BerthListView[]
       screen BerthListScreen
         data BerthListView[] via query ListBerths
-        action BookBerth
-          navigate to BookBerthScreen
+        section bookBerthInput
+          title "Book a berth"
+          on click
+            navigate to BookBerthScreen
 
   form BookBerthForm for BookBerth
     field berthId label "Berth"
@@ -58,8 +63,11 @@ payload field the event never declares, so the event payload line matters.
 - `identifier` marks which property names the event source. It neither generates the value nor
   fills it. Here the user supplies the berth. Other sources are a prefill (`populate via query`
   or `populate from item` when the user chose a berth on the previous screen) or a modeled
-  selection. `generated identifier` is ESM v7: it binds on standalone Screenplay 4.68.0 and reports `PLAY0268`
-  on the 4.66.0 bundled in `cratis` 3.28.x; Stage 4.24.2 does not render it.
-- The form is found through `action BookBerth`; neither screen names it.
-- Compiled with `screenplay` 4.68.0 `--warnaserror` and `cratis screenplay validate
-  --warnings-as-errors` 3.28.2.
+  selection. `generated identifier` is ESM v7: it binds on standalone Screenplay 4.68.0 and later and on the current
+  cratis bundle (`PLAY0268` on the 4.66.0 bundled in `cratis` 3.28.x); Stage 4.73.1 refuses it
+  (`STAGE-ESM-028`).
+- The form is found through the input screen's `action BookBerth`; neither screen names it. A click
+  opens `BookBerthScreen` before the command runs; the action's `navigate to` is where the user
+  lands after success, never a route to the command's own input screen (Screenplay `screens.md`).
+- Compiled with `screenplay` 4.139.0 `--warnaserror`; `cratis screenplay validate --executable` 3.53.1
+  reports it valid (bundled 4.137.0).