Release v7.9.0

Summary

The Screenplay event-modeling skills and agents now adopt guidance from agentic-engineer by Martin Dilger and Nebulit GmbH closely, with their agreement, adapted to Screenplay and Cratis.

Added

• A completeness self-check ("do not cut corners to save tokens or effort"), a per-turn protocol and fuller phase-transition and reasoning-note guidance in cratis-screenplay-modeling-lifecycle (#497)
• The business-question lens and an automation-chain audit as review checks in cratis-screenplay-model-review, which now has 89 checks (#497)
• Build prompts and conformance checklists for hand-written slices in cratis-application-slice-conformance, including a check that the running application really subscribes each reactor (#497)
• A candidate inventory and UI-walk progress reporting in cratis-screenplay-legacy-extraction (#497)
• Good and bad scenario forms, with a runnable invoicing example and a design-mode companion that declares its stored-state rules, in cratis-screenplay-scenario-coverage (#497)
• An Acknowledgments section in the README crediting Martin Dilger, Nebulit GmbH and TrogonStack (#497)

Changed

• Discovery, slice design, streams, automations and translations, render and gap-fill, and toolchain guidance follow Martin Dilger and Nebulit GmbH's wording, checklists and output formats more closely, for example storylines, events with no fields, per-record translation contracts and stream decision trees (#497)
• The Screenplay Modeler, Screenplay Reviewer, Screenplay Renderer and Slice Implementer agents point to the adopted per-run protocol, business-question lens and delivery discipline (#497)
• Each affected skill's provenance and LICENSE now credit Martin Dilger and Nebulit GmbH, with their agreement, alongside the existing TrogonStack MIT notices (#497)