kibo-template-viper

The first-party Kibo template. From a DSM model, it produces C++ surfaces for Viper C++ and a typed Python package for dsviper — the two sides of the Viper runtime.

Place in the ecosystem

• Depends on — Kibo, DSM, the Viper C++ runtime.

• Consumed by — kibo invocations targeting the Viper runtime.

• Source repository — digital-substrate/kibo-template-viper.

• Distribution — bundled in the DevKit ZIP

What it produces

Two surfaces from one DSM model:

• C++ surfaces — headers and implementation files for type definitions, attachments, persistence, serialization, and function pools, designed to link against the Viper C++ engine (commercial).

• A typed Python package — classes, functions, and type hints mirroring the model. Importable as import <namespace> and usable through the dsviper runtime (PyPI).

Each generated surface is the output of one or more templated features. The catalogue is in Templated Features.

The Dual Reality

The code produced by kibo-template-viper is not a runtime in its own right — it is a static adapter layer over Viper’s dynamic API. Both APIs are strongly typed ; the type catalog, built from DSM at runtime, is the single source of truth. The generated adapter gives developers and IDEs a static, idiomatic surface (typed classes, STL in C++, type hints in Python) that delegates to the dynamic runtime transparently. The static surface does not add type safety — it adds editor ergonomics on top of safety that already lives in the runtime.

This split is the Dual Reality pattern :

• Developer Reality — statically typed at edit/compile time, what the IDE sees.

• Runtime Reality — dynamically typed via the metadata catalog, what the engine actually manipulates.

Both realities coexist for the same data; the generated adapters translate between them, sharing one engine — no parallel implementation, no C++/Python divergence. Applications that do not need the typed comfort can bypass the static side and work directly against the dynamic runtime (see static API / dynamic API) — that is what generic tools like cdbe.py do.

Topics

• Templated Features
  • C++ Templates
  • Python Templates
  • Picking a feature set
• Python Wheels
  • Quick Start
  • Package Structure
  • Distribution
  • Version Management

Status

Part of DevKit 1.2.x (LTS, feature-locked). For Kibo’s CLI, the generic template format, and how to write your own template targeting another runtime, see the Kibo section.