@cratis/pi 7.21.0 → 7.22.0

package/package/corpus/skills/cratis-screenplay-model-review/references/checklist.md

@@ -74,7 +74,7 @@ the format in `worked-example.md`.
 
 | Id | Check | Evidence | Tier |
 |---|---|---|---|
-| A1 | Every read-model field maps from an event field, `$eventSourceId`, `$eventContext`, a literal, or a documented derivation (`count`, `add ... by`) | projection blocks; AutoMap name matches (on by default) | D (a declared field that nothing maps is reported by no tool; a mapping to an undeclared property fails binding at V3, PLAY0273, tier C when executable) |
+| A1 | Every read-model field maps from an event field, `$eventSourceId`, `$eventContext`, a literal, or a documented derivation (`count`, `add ... by`) | projection blocks; AutoMap name matches (on by default) | D (opt-in PLAY0534 detects known missing origins; unknown coverage still needs a walk. Absent mapping targets warn at V1 with PLAY0514, tier C for source validity) |
 | A2 | Every event field comes from a command input, `$context` value, literal, capture field or reaction mapping | `produces` mappings, capture `map`, reaction `invokes` | C for undeclared names (warning), D for meaning |
 | A3 | Every command input is supplied by someone: screen form field, reaction mapping, capture, or a caller who knows it | screens, forms, reactions | D |
 | A4 | Every command input ends in an event field or is used by a rule | `produces`, `validate`, `require` | Q (an unused input is often a missed fact) |
@@ -100,7 +100,7 @@ the format in `worked-example.md`.
 | C2 | A fact, not a request, a running calculation, a UI action or a technical step | names and payloads | D |
 | C3 | No speculative or pending events ("MayBe...", "...Pending") where a decision is meant | names | Q |
 | C4 | No nullable or optional event property by default: a required value, or a separate event for the second situation (CHR0012; a `null` in a specification value is PLAY0350); a kept `optional` has its reason in the `description` | `optional` members | D |
-| C5 | Sensitive members marked (`@pii`/`@sensitive`) and not mixed across data subjects; operational secrets are `@sensitive`, not `@pii`; no `@pii` or `@sensitive` on the event-source id (no such identifier concept; use a surrogate id; Chronicle CHR0034 and CHR0052, and Screenplay PLAY0515 from 4.69.1 for `@pii` and 4.84.1 for `@sensitive`, only on tools at those versions - the pinned tools do not report it); one data subject per event or stream | concepts and members | D; removal to pass tools = critical |
+| C5 | Sensitive members marked (`@pii`/`@sensitive`) and not mixed across data subjects; operational secrets are `@sensitive`, not `@pii`; no `@pii` or `@sensitive` on the event-source id (no such identifier concept; use a surrogate id; Chronicle CHR0034 and CHR0052, and Screenplay PLAY0515 from 4.69.1 for `@pii` and 4.84.1 for `@sensitive`, only on tools at those versions - current pins report it as a source error, including typed destinations and route-id parts); one data subject per event or stream | concepts and members | D; removal to pass tools = critical |
 | C6 | No secret or bearer values as facts (tokens, magic links, signed URLs): record a keyed hash or a reference | event payloads | D |
 | C7 | No two events of one kind of event source mean the same thing (a shared property set is a signal, not a defect: distinct facts may share a shape and historical generations are exempt; major only when meaning is lost, a rule bypassed or consumers ambiguous); one fact is not recorded twice (aggregate event plus per-item fan-out) | event declarations side by side | D |
 | C8 | Corrections are new facts that name what they correct; redaction and crypto-erasure are compliance tools, not corrections | event names | D |
@@ -131,6 +131,15 @@ the format in `worked-example.md`.
 | E9 | A decided refusal is a `then error` or `then denied` specification (permanent behavior), never a lingering open question; only a genuinely unanswered question counts as open, and resolving it means answering it | specs, `STATE.md` questions, descriptions | D |
 
 ## Phase 8. Flow and lifecycle (13 checks, F1-F13)
+
+Review PLAY0516 backward-flow and PLAY0517 sibling-cycle findings as
+**presentation-only** information: they do not fail `--warnaserror`, change
+execution or prove a business deadlock. An ordering root is required; without
+one the folder check is skipped. A real F6 defect needs a named domain
+consequence, not just the diagram's cycle. Suggested graph order and safe
+PLAY0516 moves preserve executable bytes/identities. Source: Screenplay
+`v4.125.0:Documentation/screenplay/imports.md`, `diagnostics.md`.
+
 | Id | Check | Evidence | Tier |
 |---|---|---|---|
 | F1 | Each entity has a first fact, a state-transition table (including refused transitions) and terminal states | slices, constraints, `description` | D |

package/package/corpus/skills/cratis-screenplay-read-surface/SKILL.md

@@ -231,6 +231,16 @@ block or a named `behavior` attached with `uses`. See
 `cratis-screenplay-ui-composition`. Screens and interactions are deferred from the
 executable model (information `PLAY0269`); they never block binding.
 
+## Guarded screen actions
+
+One labeled user decision can choose among commands using ordered
+`when item.<path> ... execute <Command>` alternatives, with optional last
+`otherwise execute` or `otherwise hidden`. First match wins; no item or no match
+without an execute fallback hides it. Inputs use typed `with ... from ...`
+bindings; the selected command still enforces its own gate/rules/constraints.
+Read [guarded actions](references/guarded-actions.md) for the checked model,
+missing-versus-null behavior and source-versus-renderer boundary.
+
 ## How a bare name resolves
 
 **Inside out:** the slice, then the enclosing feature, then the module, then the

package/package/corpus/skills/cratis-screenplay-read-surface/references/guarded-actions.md

@@ -0,0 +1,79 @@
+<!-- Copyright (c) Cratis. All rights reserved. -->
+<!-- Licensed under the MIT license. See LICENSE file in the project root for full license information. -->
+
+# Guarded screen actions
+
+## Contents
+
+- State one user decision
+- Select and bind
+- Guard boundaries
+
+## State one user decision
+
+A label-headed `action "<text>"` (or `$strings.<key>`) can choose a command
+from the displayed item's state. The label replaces a plain action's `label`
+child. This complete model compiled without warnings on Screenplay 4.125.0;
+it is authoring evidence, not a rendered interaction test:
+
+```screenplay
+module Tasks
+  feature Work
+    slice StateChange CloseTask
+      command CloseTask
+        taskId String identifier
+        produces event TaskClosed
+    slice StateView Details
+      readmodel TaskDetails
+        taskId String
+        status String
+      query TaskById => TaskDetails optional
+        by taskId String
+      screen Details
+        data TaskDetails via query TaskById by taskId
+        action "Close"
+          when item.status == "open" execute CloseTask
+            with taskId from item.taskId
+          otherwise hidden
+```
+
+## Select and bind
+
+`item` is the nearest container's data item, or selected element for collection
+data. Sibling data in the action's container takes precedence over outer data
+regardless of text order; ambiguous/missing data warns rather than guessing.
+Without an item (loading/no selection), always hide the action, even if it has an
+execute fallback.
+
+Check alternatives in authored order whenever data changes; **first match wins**.
+At least one when is required. Last optional `otherwise execute <Command>` is
+a fallback; `otherwise hidden` or omission hides unmatched actions. Intentional
+overlap is valid; only provable shadowing warns. Inputs resolve from explicit
+`with <property> from <binding>`, matching subject fields, declared command form,
+then renderer input. Terminal collection fields may fill collection inputs;
+traversing through a collection is not supported.
+
+A single `navigate to` runs after the chosen command succeeds, not to open that
+command's input screen. Open a form/dialog first when it needs separate input.
+A click runs the choice shown; if a click-time check changes it, refresh instead
+of executing the newly selected command.
+
+## Guard boundaries
+
+Conditions compare item paths to literals, not other paths, route/screen state,
+`$context` or `$env`. Equality/inequality, numeric ordering, string contains/starts
+with and grouped and/or are supported; and binds tighter than or. Collections
+cannot be condition operands. Missing/null fields compare false, including !=
+to a non-null literal; == null matches an explicit null, not a missing member.
+Type mismatches compare false. Editors validate syntax/shape, not runtime choices.
+
+The guard controls an offer, never server permission. The selected command still
+runs authorization, validation and constraints; denial never falls through to a
+second command. Screens remain outside portable ESM execution. A downstream
+renderer must explicitly admit guarded selection or report unsupported behavior,
+not approximate it as a plain action. CLI 3.41.0's bundled screen renderer reports
+unsupported guarded actions with STAGE-SCENE-ACTION-001; source compilation alone
+cannot establish usable controls.
+
+Authority: Screenplay `v4.125.0:Documentation/screenplay/screens.md`;
+CLI `v3.41.0:Documentation/reference/screenplay.md` for renderer disposition.

package/package/corpus/skills/cratis-screenplay-render-and-gap-fill/SKILL.md

@@ -25,9 +25,9 @@ Version pins live once in `cratis-screenplay-toolchain` `references/versions.md`
 
 | Source | Pin | Used for |
 | --- | --- | --- |
-| cratis CLI | `3.40.7` (`v3.40.7`, `af9f18e`) | `cratis render` options, exit codes, publication receipt; tag pins Stage 4.51.1 and Screenplay 4.114.0, with `cratis/stage:4.51.1` as the runtime image. UI-profile rejection and render-then-recover fixes are pending in Cratis/cli#301. Reading baselines: 3.40.3, 3.39.0 and 3.28.3 (`8b43fef`) |
-| Stage | `4.51.1` (CLI default; latest package 4.51.3) | admission (`STAGE-ESM-*`, and `PLAY0268` render refusals surfaced by the CLI), ownership manifest, `Customizations/` seams, Debug-only specifications, fail-closed guarded actions (`STAGE-SCENE-ACTION-001`) and interactions (`STAGE-SCENE-INTERACTION-001`). Older baselines: 4.49.5, 4.43.0 and `v4.24.2` (`32dcac4`) |
-| Screenplay | bundled `4.114.0`; latest language package `4.122.0` | standalone compiler for V1 to V3; binding codes `PLAY0268` and the code-attachment rules, read at `v4.66.0`; claims made through `cratis` are 4.114.0 claims |
+| cratis CLI | `3.41.0` (`v3.41.0`) | `cratis render` options, exit codes, publication receipt, UI-profile rejection (`CLI-RENDER-007` to `CLI-RENDER-010`); tag pins Stage 4.51.3, Screenplay 4.114.0 and `Cratis.Arc.Screenplay` 22.54.0, with `cratis/stage:4.51.3` as the runtime image. UI-profile rejection and render-then-recover shipped in 3.41.0 (Cratis/cli#301). The screens release vector was verified on 3.40.7 (`af9f18e`) with `cratis/stage:4.51.1`. Reading baselines: 3.40.3, 3.39.0 and 3.28.3 (`8b43fef`) |
+| Stage | `4.51.3` (CLI 3.41.0 default) | admission (`STAGE-ESM-*`, and `PLAY0268` render refusals surfaced by the CLI), ownership manifest, `Customizations/` seams, Debug-only specifications, fail-closed guarded actions (`STAGE-SCENE-ACTION-001`) and interactions (`STAGE-SCENE-INTERACTION-001`). Older baselines: 4.49.5, 4.43.0 and `v4.24.2` (`32dcac4`) |
+| Screenplay | bundled `4.114.0`; standalone `4.125.0` | standalone compiler for V1 to V3; binding codes `PLAY0268` and the code-attachment rules, read at `v4.66.0`; claims made through `cratis` are 4.114.0 claims |
 | Rendered apps | Arc `22.25.0`, Chronicle `19.8.1`, Scene packages | `[ProtectedDecision]` (Arc 22.39.0 and later) is not available in code written into one; verify the generated package set before claiming exact Scene 4.14.0 runtime parity |
 
 Facts were read at those tags and `cratis render` was run at 3.28.2 and 3.28.3 for the worked example
@@ -68,13 +68,15 @@ assume an answer to 4, and record it as `blocked` for the user.
 Evolved events (a model that selects ESM v4) are gap-fill too: Stage 4.24.2 admits the v4 model but refuses any evolved event and its dependent scope with `STAGE-ESM-026`, because it cannot render Chronicle event-type migrations yet (Stage#204). Events at their initial revision still render. Hand-write the evolved events and their migrations per `cratis-chronicle-event-type-migration`, with the model as the contract. The cratis CLI bundles Stage 4.24.2 since 3.28.3 and reports an evolved event with `STAGE-ESM-026` (probed); 3.28.2 bundled Stage 4.24.1 (3.28.1 and earlier: 4.24.0) and refused a v4 model with `STAGE-ESM-016`.
 
 Generated command values and `returns` responses (ESM v7) bind and run in the reference runner on standalone
-Screenplay 4.68.0 but are not rendered: the cratis CLI 3.28.3 bundles 4.66.0 and reports `PLAY0268`, and Stage
-4.24.2 refuses ESM v7 with `STAGE-ESM-016` (tracked in Stage#201). A command that uses them is case A or C: keep
+Screenplay 4.68.0 and later but are not rendered: CLI 3.41.0 admits ESM v7 to Stage planning, and Stage 4.51.3
+refuses generated values and responses with `STAGE-ESM-028`/`STAGE-ESM-029` (`cratis-screenplay-toolchain`
+`references/versions.md`). A command that uses them is case A or C: keep
 the model, hand-write the command and its response with the model (and its fixtures and `then returns`
 specifications) as the contract, and report V3 from the standalone tool and V5 as `not run: no render`.
 
-Models with event sources, streams or command routes are case A or C today: they cannot
-bind or render (`PLAY0268`: not admitted by any supported ESM version, so also `STAGE-ESM-016`), so hand-write the code with the model as the
+Models with event sources, streams or command routes bind and reference-execute as ESM v8 on standalone
+Screenplay 4.125.0, so report V3 and V4 from that tool. They are still case A or C for rendering: CLI 3.41.0
+refuses models above ESM v7 with `CLI-RENDER-004`, so hand-write the code with the model as the
 contract (`cratis-screenplay-toolchain` `references/sources-and-streams.md`).
 
 Two rules hold in every case: a customization never makes a rejected model renderable, and

package/package/corpus/skills/cratis-screenplay-render-and-gap-fill/references/gap-fill-handoff.md

@@ -75,9 +75,12 @@ gap-fill): the registration goes through the seams in `Customizations/Program.cs
 not bind a modeled reaction or capture. Case C (no generated application): there is no generated
 `Program.cs` or partial hook, so the registration follows the hand-written application's own
 composition conventions (`rules/vertical-slices.md`, `cratis-chronicle-reactor`).
-A reaction invokes a command with no caller (Screenplay 4.68.0 has no caller syntax), so the
-code needs a trusted path (`[ExecuteCommandsAsSystem]`, Arc 20.56.0 and later, in a rendered
-application's Arc 22.25.0 too); never invent `runs as`.
+Declare the trusted actor with reaction `runs as system role "<Role>"`
+(standalone 4.125.0, ESM v10); `given caller` is not an invocation actor.
+CLI 3.41.0's older compiler rejects the syntax, and Stage 4.51.3's planner admits
+through v7 only. Gap-fill retains the declared role and uses Arc
+`[ExecuteCommandsAsSystem]` for returned commands. Checked model and diagnostics:
+captures-and-reactions `references/invocation-identity.md`.
 
 ## Re-delivery
 The work list is the full in-scope contract delta since the ledger's last verified source

package/package/corpus/skills/cratis-screenplay-render-and-gap-fill/verification.json

@@ -37,14 +37,14 @@
     },
     {
       "kind": "skill-contains",
-      "value": "case A or C today"
+      "value": "case A or C for rendering"
     },
     {
       "kind": "skill-contains",
       "value": "STAGE-ESM-026"
     },
     { "kind": "skill-contains", "value": "Generated command values and `returns` responses (ESM v7) bind and run" },
-    { "kind": "skill-contains", "value": "tracked in Stage#201" },
+    { "kind": "skill-contains", "value": "STAGE-ESM-028`/`STAGE-ESM-029" },
     { "kind": "skill-contains", "value": "authored-UI parity" },
     { "kind": "skill-contains", "value": "responsible tool version" }
   ]

package/package/corpus/skills/cratis-screenplay-scenario-coverage/SKILL.md

@@ -183,8 +183,9 @@ scope is reported as explicitly incomplete); V1 passes with the tool named in
   populated-then-removed spec.
 - V1 passes (standalone `screenplay` required when cascades are specified); report tool and
   version.
-- V4 is "not run: no route" unless a Screenplay reference route was actually executed (Stage's spec runner is target-engine evidence, never V4); V1/V3 do not prove specs
-  pass. V5 is "not run" unless rendering was attempted; report admission, publication, build and
+- Run V4 through standalone `screenplay test` or MCP `run-specifications`;
+  if not executed, report the actual host/input reason, not a blanket no-route
+  claim. Stage's runner is target-engine evidence, never V4; V1/V3 do not prove pass. V5 is "not run" unless rendering was attempted; report admission, publication, build and
   tests separately.
 - Match guarantees to evidence: `.play` outcomes, in-process Chronicle scenarios and real-kernel
   tests are not interchangeable (`references/chronicle-verification.md`).

package/package/corpus/skills/cratis-screenplay-scenario-coverage/references/spec-forms-by-mode.md

@@ -3,7 +3,11 @@
 The subsets themselves (what binds, what renders) are owned by `cratis-screenplay-toolchain`
 (its executable and renderable subset references); version facts by its `references/versions.md`. This file
 says what that means for writing specs. A form outside the chosen mode is never deleted from a
-design model: keep it and record the gap.
+design model: keep it and record the gap. The detailed Stage 4.24.x rows below
+retain historical admission evidence; current pins are Screenplay 4.125.0 / CLI
+3.41.0. Run actual target admission instead of assuming old restrictions carry over.
+Typed example/override forms: `cratis-screenplay-specifications`
+`references/examples-and-routes.md`.
 
 | Form | design | executable | renderable |
 |---|---|---|---|
@@ -19,6 +23,10 @@ design model: keep it and record the gap.
 | Command spec listing reaction cascade events | yes (v6 semantics) | false PLAY0285 on a cratis before 3.28.2 bundle only; see SKILL.md "Version skew" | no (no reactions) |
 | `generated <name> = <value>` under `when`, `then returns` (ESM v7) | yes | yes on standalone Screenplay 4.68.0 or later; not on the 4.66.0 bundled in cratis 3.28.x (`PLAY0268`) | no (Stage does not render ESM v7) |
 | `then events in any order` | yes | yes | yes |
+| Event `stream` / `no stream`, typed `for` and stream ids | yes | ESM v8 at 4.125.0 | CLI 3.41.0 refuses models above v7 before planning |
+| Typed examples and overrides | yes | expanded and type-checked before binding; required fields remain required | depends on expanded contract, not fixture reuse |
+| `given operation ... fails`, `then operation`, `then compensated` | yes | unadmitted PLAY0268 | gap-fill |
+| Refusal branches, `when redelivered`, `then no events` | yes | unadmitted PLAY0268 | gap-fill |
 | Composite (JSON) values in specs | yes | yes | no |
 | Rules over dates or `today`, nested paths | spec them in design | rule does not bind: `recorded` | `recorded` |
 | Bodied rules, code policies | spec them | spec reported **unsupported**, never passed | no |
@@ -36,7 +44,10 @@ bundle is available). Mark each such spec in the coverage matrix as
   the projection's effective key (`cratis-screenplay-slice-design` `references/read-model-design.md`
   decision table). View specs assert through a keyed query.
 - Run V3 (binding-ready, per `cratis-screenplay-toolchain`); binding-only codes (PLAY0273/0350/0352/0388/0389) appear only there.
-- V4 (reference specs run) needs a runner route; otherwise report "V4 not run: no route". Unsupported is not passed.
+- V4 runs through `screenplay test` or MCP `run-specifications`; record expected,
+  discovered/selected and passed/failed/unsupported counts. Unbound/unsupported
+  is not passed. Authoring-only refusal/redelivery/operation specs remain recorded
+  contracts; test them at the owning target after authorized gap-fill.
 
 ## Renderable mode
 - Specs use given events, a command `when`, and then events / readmodel / query / error /

package/package/corpus/skills/cratis-screenplay-slice-design/SKILL.md

@@ -65,8 +65,9 @@ report and the slice `description`, and continue.
    origin must be **reachable** (some screen, form, reaction or integration leads to the
    command) and **authorized** (the persona's policies satisfy every gate on module, feature
    and command). A command may have several origins; each must be real. A reaction's `invokes`
-   has no caller, so a gated command it invokes is refused: design the trusted path, never
-   delete the gate.
+   has no caller unless the reaction declares `runs as system role "<Role>"` (ESM v10), so a
+   gated command it invokes is refused: declare that trusted path with the role the gate
+   admits, never delete the gate.
 3. **Design the command** (`references/command-inventory.md`). Imperative business name for
    **one** decision (`ReassignLocker`, `CorrectLockerVolume`). A form verb (`Update`, `Edit`,
    `Save`, `Set`, `Change`, `Manage` + noun) means several decisions share a form: ask which

package/package/corpus/skills/cratis-screenplay-slice-design/references/command-inventory.md

@@ -118,8 +118,9 @@ not been asked the refusal questions.
 - Copying state into the command so `require` can read it, instead of `reads <View>`.
 - An origin that is neither reachable (no screen, form or reaction leads there) nor authorized
   for the persona who should act.
-- A reaction invoking a gated command: the invocation has no caller and is refused; design the
-  trusted path (`cratis-screenplay-automations-and-translations`), never delete the gate.
+- A reaction invoking a gated command without `runs as system role "<Role>"`: the invocation has
+  no caller and is refused; declare the trusted path with the role the gate admits
+  (`cratis-screenplay-automations-and-translations`), never delete the gate.
 - Refusals recorded as prose in the `description` with no layer and no spec row.
 - Personal data or secrets as command input: values are by default recorded in the causation chain, so treat them as permanent
   (retention can be configured in newer Chronicle; verify it, never audit secrets).

package/package/corpus/skills/cratis-screenplay-slice-design/references/slicing.md

@@ -62,6 +62,46 @@ For `worked-example.md`:
 | LockerBoard | LockerInstalled, LockerAssigned | InstallLocker, AssignLocker | |
 | RequestProgress | LockerRequested, LockerAssigned | RequestLocker, AssignLocker | |
 
+### Declare container dependencies
+
+At Screenplay 4.125.0, modules and features may declare `depends on <Name>`, one
+target per line. Use a bare sibling name or a qualified module/feature address.
+Bare names resolve inside out: siblings, ancestor siblings, then root modules;
+qualified names match an unambiguous trailing address. Ambiguity is `PLAY0198`.
+Self, ancestor, descendant and unresolved targets warn with `PLAY0554`; repeats
+keep the first resolved target and warn with `PLAY0555`.
+
+Declarations accumulate on the same owner in ordinal file-path order, independent
+of board import order. Layout expansion writes them only in the owner's file.
+They are authoring metadata: no execution rule, implementation order, identity
+or executable-model revision change.
+
+Once a container declares dependencies, its own inventory must cover counted
+explicit references leaving it, including descendants. Parent and child lists
+are independent. Module targets cover all their features; feature targets cover
+their descendants. `usesFactsFrom`, `reactsTo`, `decidesFrom`, `asks` and `shows`
+count; specification-only and outside-context edges do not. `PLAY0552` warns on
+uncovered producer modules; `PLAY0553` marks unused declarations and `PLAY0556`
+mutual declarations as information. Ambiguous ownership is provisional, not proof
+of an undeclared dependency. Inspect MCP `dependency-graph`'s `declarations` view.
+
+Example: Payroll declares Timesheets; its Handover feature declares
+Timesheets.Approval and sibling Runs. These are dependency inventories, not
+actual references or execution dependencies:
+
+```screenplay
+module Payroll
+  depends on Timesheets
+  feature Handover
+    depends on Timesheets.Approval
+    depends on Runs
+  feature Runs
+module Timesheets
+  feature Approval
+```
+
+Source: Screenplay `v4.125.0:Documentation/screenplay/slices.md` and `folders.md`.
+
 ## State-transition table (per entity)
 
 Rows are states (named after the fact that put the entity there); columns are commands. Each

package/package/corpus/skills/cratis-screenplay-specifications/SKILL.md

@@ -37,15 +37,12 @@ The update follows `commands.md`, `events.md`, `types.md` and
 `specifications.md` at that main commit (after v4.52.0). This pass compiles the
 changed examples; it does not rerun the reference runner.
 
-**Which tool says what.** The standalone `screenplay` tool 4.68.0 admits ESM v1-v7.
-The `cratis` CLI 3.28.2 and 3.28.3 bundle Screenplay 4.66.0 (ESM v1-v6: generated fixtures and `then returns` report `PLAY0268`). Before `cratis` 3.28.2 the
-bundle was Screenplay 4.60.1, which admitted ESM v1-v5 only and reported a false
-`PLAY0285` on reaction cascades (see "Version skew" in [references/reactions-and-cascades.md](references/reactions-and-cascades.md)). Facts below marked **ESM v6**
-hold for Screenplay 4.61.0 and later: the standalone tool and `cratis` 3.28.2 or later. The full table is in `cratis-screenplay-toolchain`
-(`references/versions.md`; read it before choosing a compiler or interpreting version-dependent binding diagnostics). Neither tool runs specifications: `screenplay` and
-`cratis screenplay validate` parse and check consistency, the MCP server also binds,
-and spec outcomes come only from the reference runner as a library, Stage's
-specification runner or a rendered application's tests.
+Current specification syntax follows Screenplay v4.125.0 `specifications.md`;
+older probes in the table remain historical. Pins: `cratis-screenplay-toolchain`
+`references/versions.md`. Compile commands check source, not reference execution.
+Use `screenplay test` or MCP `run-specifications` for V4; Stage/target tests are
+separate evidence. Unsupported/unbound is never a pass. Typed examples and
+route assertions are in [examples and routes](references/examples-and-routes.md).
 
 Checked against the Screenplay repository at tag `v4.31.0` (commit `355dffb`):
 `Documentation/screenplay/{specifications,policies,constraints,readmodels,diagnostics}.md`.
@@ -95,9 +92,8 @@ target has one normalized behavior to match.
   rejects, the earlier accepted facts and their projections remain.
 
 Unsupported is not passed. Report it as "needs a target", not as green. "Specified"
-and "bound" are never "passing": no command-line tool runs these specifications, so a
-pass is claimed only from the reference runner, Stage's specification runner or a
-rendered application's tests, and named as such.
+and "bound" are never "passing": run the reference CLI/MCP route or the owning
+target tests, and name the engine, version and expected/discovered selection.
 
 ## The vocabulary
 
@@ -184,7 +180,7 @@ Read [references/worked-example.md](references/worked-example.md) when writing o
 
 ## Generated fixtures and return expectations (ESM v7)
 
-Standalone `screenplay` 4.68.0 or later binds and runs generated fixtures (`generated <name> = <value>`, `for` beneath `when`) and `then returns`; the cratis CLI 3.28.3 does not. If execution reaches generation without every generated value supplied, the run is unsupported and never passes.
+Current standalone and bundled compilers bind generated fixtures (`generated <name> = <value>`, `for` beneath `when`) and `then returns` as ESM v7. If execution reaches generation without every generated value supplied, the run is unsupported and never passes.
 Read [references/generated-fixtures.md](references/generated-fixtures.md) when writing generated-value fixtures or scalar and record return expectations for a command.
 
 ## Actions beyond commands
@@ -197,6 +193,18 @@ Read [references/actions-beyond-commands.md](references/actions-beyond-commands.
 From ESM v6 every action sets reactions off and `then` events compare every new fact, the action's and the reactions'. A specification whose action is a command goes in the slice that declares the command. A false `PLAY0285` on a cascade under `cratis` before 3.28.2 is tool skew, not a modeling error: never delete the `then` line to silence it.
 Read [references/reactions-and-cascades.md](references/reactions-and-cascades.md) when a specification's expected events include what a reaction appends, when a cascade, clock tick or invoked command is involved, or when `cratis` reports `PLAY0285` on a model that `screenplay` accepts (it holds the "Version skew: cascades and the false `PLAY0285` (cratis before 3.28.2)" section).
 
+## Reusable fixtures, routes and recovery intent
+
+Use typed `example` declarations for repeated event/command/read-model values;
+step overrides keep the action/outcome visible and carry authored/example/override
+provenance. Event `stream`/`no stream` assertions bind as ESM v8; typed `for` and
+stream ids are not guessed. Read [examples and routes](references/examples-and-routes.md).
+
+`given operation ... fails`, `then operation` and `then compensated` describe
+future external-work behavior. Reaction refusal branches, `when redelivered`
+and `then no events` are also authoring-only: compile them, but report their
+PLAY0268 binding refusal instead of an executed result. Preserve the contract.
+
 ## Rejections and denials say different things
 
 `then error "<message>"` says **rejected, for this reason** — pin it when the
@@ -252,9 +260,9 @@ For view specifications:
 
 ## Verify
 
-- [ ] `screenplay <model> --warnaserror` reports zero errors and zero warnings
-      (standalone 4.68.0 or `cratis` 3.28.2 or later, bundling 4.66.0; under `cratis` before 3.28.2 expect the false
-      `PLAY0285` on reaction cascades and record that it is skew).
+- [ ] Validate the complete root with warnings as errors, naming tool/version.
+      Current standalone pin: 4.125.0. Run specification examples through
+      `screenplay test`; expected unadmitted syntax reports unbound, not passed.
 - [ ] Executable diagnostics are clean too: `PLAY0350`, `PLAY0352`, `PLAY0388`,
       `PLAY0389` and `PLAY0273` are only reported at binding.
 - [ ] Every slice has at least one boundary specification.

package/package/corpus/skills/cratis-screenplay-specifications/references/examples-and-routes.md

@@ -0,0 +1,103 @@
+<!-- Copyright (c) Cratis. All rights reserved. -->
+<!-- Licensed under the MIT license. See LICENSE file in the project root for full license information. -->
+
+# Typed examples, overrides and event routes
+
+## Contents
+
+- Typed examples
+- Effective values
+- Event routes
+- Operation and recovery intent
+
+## Typed examples
+
+`example <Name> : <EventOrCommandOrReadModel>` declares a possibly partial typed
+fixture at document/module/feature/slice scope. Its underlying type resolves in
+the declaration scope, not the use scope. It can state properties, description,
+`for` and command generated fixtures, not callers/clocks or a whole scenario.
+Examples share the type namespace and use the current event generation. They
+cannot inherit from examples or serve as query results/returns/trigger/capture
+fixtures. They are specification data, not operational `seed` events.
+
+This complete example compiled and its two reference scenarios passed on 4.125.0:
+
+```screenplay
+concept ProjectId : Uuid
+module Projects
+  feature Registration
+    example DefaultProject : RegisterProject
+      name = "Default"
+    slice StateChange Register
+      command RegisterProject
+        projectId ProjectId identifier
+        name String
+        produces event ProjectRegistered
+          name String = name
+      specification RegisteringTheDefault
+        when DefaultProject
+          projectId = "3fa85f64-5717-4562-b3fc-2c963f66afa6"
+        then ProjectRegistered
+          for "3fa85f64-5717-4562-b3fc-2c963f66afa6"
+          name = "Default"
+      specification OverridingTheName
+        when DefaultProject name = "Override"
+          projectId = "3fa85f64-5717-4562-b3fc-2c963f66afa6"
+        then ProjectRegistered
+          for "3fa85f64-5717-4562-b3fc-2c963f66afa6"
+          name = "Override"
+```
+
+## Effective values
+
+An example occupies the ordinary name slot in given/event, given readmodel,
+when command, when append, then event or then readmodel. One assignment may be
+inline after the name; more stay indented. There is **no with keyword**.
+Step values override example values by property name; objects/lists replace
+whole values, never recursively merge. Indented `for` replaces its destination;
+generated fixtures merge by name. A new step-only value is authored, not an
+override. Each step/declaration assigns a property once (PLAY0519); malformed
+headers are PLAY0518.
+
+After expansion, required given-event/read-model, command, append and expected
+event fields must be complete (binding PLAY0524); examples can remain partial.
+Read-model assertions retain subset matching unless `exactly` is authored.
+Even unused example values or values overridden at every use are type-checked.
+No default fixture values are invented. A command action must belong to the
+specification's own slice even when a qualified example resolves elsewhere.
+
+MCP `find-fixtures` reports effective authored/example/override origins and
+replaced values. Source-bound runner failures retain that provenance; canonical
+ESM contains merged values only, equivalent to hand-expansion. ESM-only runners
+cannot reconstruct source origins.
+
+## Event routes
+
+Event occurrences in `given`, `when append` and `then` accept `stream Source.Stream`
+with nested scalar `streamId = <literal>` or a complete composite part block.
+These select ESM v8 and reference-execute at 4.125.0; the complete checked model
+is in toolchain `references/sources-and-streams.md`. Command actions take their
+route from the command declaration, never the fixture.
+
+Routed history/appends need typed concrete `for`; routed expectations may omit
+it. `then ... no stream` expects no route; no route in a then is a wildcard.
+`no stream` is invalid on given/appends. Event examples can supply routes, but
+steps replace the entire route, independently of `for`; an inherited route
+cannot be changed back to a wildcard. Literal compatibility/route contradiction
+checks use resolved source/stream types, not guessed text (PLAY0547–PLAY0551).
+CLI 3.41.0 rendering rejects versions above v7 before planning, even though
+source compilation and reference routing pass.
+
+## Operation and recovery intent
+
+Operation failure/request/compensation forms are authoring-only (PLAY0268):
+`given operation <Name> fails`, `then operation <Name>` with a partial input
+subset, and leaf `then compensated <Name>`. See command-surface
+`references/operations.md` for the complete checked authoring fixture.
+
+Refusal selectors, redelivery to one observer and `then no events` likewise
+remain unadmitted; they are not executed append scenarios. Forms, matching,
+acknowledgement and their whole-model refusal boundary are in
+captures-and-reactions `references/refusals-and-redelivery.md`.
+
+Authority: Screenplay `v4.125.0:Documentation/screenplay/specifications.md`.

package/package/corpus/skills/cratis-screenplay-specifications/references/reactions-and-cascades.md

@@ -42,12 +42,11 @@ module Collections
 - A specification whose action is a command goes in the slice that declares that
   command, here `SendInvoice`, not in the `Automation` slice that reacts: a command
   from another slice is `PLAY0273` "command is unresolved in its slice" at binding.
-- A command a reaction `invokes` runs through its full pipeline **with no caller**.
-  A command that needs one rejects, and that rejection ends the scenario. Each
-  invoked command is atomic; the cascade is not, so earlier accepted facts stay.
-  There is no syntax for an identity (Screenplay#383). See
-  `cratis-screenplay-automations-and-translations` and
-  `cratis-screenplay-captures-and-reactions`.
+- A reaction invocation without a declared actor has no caller and a gated
+  command denies. At 4.125.0 reaction `runs as system role "<Role>"` supplies the
+  authenticated system actor (ESM v10); `given caller` does not. Each command is
+  atomic, not the cascade. Checked actor scenario and CLI/Stage limitations:
+  captures-and-reactions `references/invocation-identity.md`.
 - The clock is UTC and exact. An `every` or `at` occurrence fires once when it is
   due after `given clock` and at or before `when clock`; `when clock` needs
   `given clock`, so the same instant twice fires nothing.

package/package/corpus/skills/cratis-screenplay-streams-and-consistency/SKILL.md

@@ -87,10 +87,10 @@ option, mark it `ASSUMED` in the slice or module `description` and in the sessio
    invariant consequences, not names or length. For Chronicle keep event-source identity,
    namespace, process-stream routing and concurrency scope apart
    (`references/chronicle-boundaries.md`). Declared `eventsource`/`stream`/`streamId` and
-   command routes can be authored and validated but not bound, run or rendered today
-   (`PLAY0268`, `STAGE-ESM-016`): keep them as intent and hand-write the routing
-   (`cratis-screenplay-toolchain` `references/sources-and-streams.md`); no supported ESM version
-   admits them. A `generated identifier` (ESM v7, standalone 4.68.0; not rendered by Stage yet) is
+   command/specification routes bind and reference-execute as ESM v8 on 4.125.0.
+   CLI 3.41.0 rendering admits through v7 only (CLI-RENDER-004); keep the routing
+   contract and gap-fill its target realization. Forms/boundaries:
+   toolchain `references/sources-and-streams.md`. A `generated identifier` (ESM v7, standalone 4.68.0; not rendered by Stage yet) is
    generated fresh on every acceptance: it is not a retry identity, and generated values give no
    idempotency, retry or deduplication guarantee. When a retry must find the same stream, the caller
    supplies the identifier before the first attempt.

package/package/corpus/skills/cratis-screenplay-streams-and-consistency/references/stream-patterns.md

@@ -205,8 +205,10 @@ Ask for each stream:
 Some event stores decide consistency by a query over tagged events instead of one stream
 (often called dynamic consistency boundaries). Chronicle's concurrency scopes can narrow a
 check to event types, source type or stream (`concurrency` block, Screenplay
-`commands.md`), and Screenplay has syntax-only event-source/stream declarations (not admitted by any supported ESM version;
-`event-sources.md`). None of this binds on either compiler (`versions.md` in `cratis-screenplay-toolchain`). Model the stream by business
+`commands.md`). Legacy concurrency metadata still fails binding (PLAY0271),
+but sources/streams/routes bind and reference-execute as ESM v8 on 4.125.0.
+Route admission does not make a view read protected. CLI 3.41.0's render boundary
+is v7; see toolchain `references/versions.md`. Model the stream by business
 identity; when a rule truly needs a cross-stream decision, record it (see
 `consistency-and-concurrency.md`) rather than inventing a construct.
 

package/package/corpus/skills/cratis-screenplay-streams-and-consistency/verification.json

@@ -12,9 +12,9 @@
     { "kind": "skill-contains", "value": "CHR0034" },
     { "kind": "skill-contains", "value": "No calculated events" },
     { "kind": "skill-contains", "value": "decision tree and red flags" },
-    { "kind": "skill-contains", "value": "not bound, run or rendered" },
+    { "kind": "skill-contains", "value": "bind and reference-execute as ESM v8" },
     { "kind": "skill-contains", "value": "STAGE-ESM-026" },
     { "kind": "skill-contains", "value": "generated values give no\n   idempotency, retry or deduplication guarantee" },
-    { "kind": "skill-contains", "value": "no supported ESM version\n   admits them" }
+    { "kind": "skill-contains", "value": "CLI-RENDER-004" }
   ]
 }