turbo-ea-capabilities

Reference Business Capability catalogue for Turbo EA, bundled as a Python package.

All catalogue data ships inside the wheel — no network access required at runtime. Designed for offline / airgapped Turbo EA deployments.

Install

pip install turbo-ea-capabilities

Usage

from turbo_ea_capabilities import (
    load_all,
    load_tree,
    get_by_id,
    get_children,
    get_subtree,
    get_ancestors,
    load_macros,
    get_macros_for_capability,
    VERSION,
    SCHEMA_VERSION,
    GENERATED_AT,
)

print(f"Catalogue {VERSION} (schema v{SCHEMA_VERSION}), built {GENERATED_AT}")
print(f"{len(load_all())} capabilities, {len(load_macros())} macro capabilities")

root = get_subtree("BC-2")
for child in root.children:
    print(f"  {child.id}  {child.name}")

# Macro navigation overlay (optional — empty list on pre-macro snapshots)
for m in load_macros():
    print(f"  {m.id}  {m.name}  ({len(m.capability_ids)} L1s)")
print(get_macros_for_capability("BC-100"))  # → [MacroCapability(id='MC-10', ...)]

API

Capability (with new optional macro_id backlink), MacroCapability, BusinessProcess, ValueStream are all frozen Pydantic v2 models — see _models.py. The macro layer is purely additive: existing Capability fields are unchanged and SCHEMA_VERSION is not bumped when only macros are added or edited.

Macro capabilities

When the L1 layer of the catalogue is too large for unaided executive consumption, the wheel ships an optional Macro Capability overlay that groups L1s into a small set of executive-level domains for navigation. The Cross-Industry baseline currently ships 9 macros covering all 41 Cross-Industry L1s; industry-specific L1s do not (yet) have a macro layer.

The overlay is:

• Optional. Older wheels without data/macro-capabilities.json keep working — load_macros() returns [].
• Additive. No existing field on Capability/BusinessProcess/ValueStream was changed. Capability gains one new optional field, macro_id: Optional[str], populated at build time on every L1 (and inherited via prefix lookup to descendants). SCHEMA_VERSION is unchanged.
• Orthogonal. Macros do not enter the BC tree, do not participate in value-stream stage links, and do not realize processes. They group L1s — that's all.

from turbo_ea_capabilities import (
    load_macros,
    get_macro,
    get_macros_for_capability,
    get_capabilities_in_macro,
)

# 1. Render Cross-Industry as 9 executive cards
for macro in load_macros():
    if macro.industry == "Cross-Industry":
        print(f"{macro.id}  {macro.name}  ({len(macro.capability_ids)} L1s)")

# 2. Breadcrumb / category label on a capability detail page
m = get_macros_for_capability("BC-100.10.20")  # walks to L1 BC-100
# → [MacroCapability(id='MC-10', name='Enterprise Governance & Risk', ...)]

# 3. Drill into a macro
caps = get_capabilities_in_macro("MC-10")
# → [Capability(id='BC-100', ...), Capability(id='BC-110', ...), ...]

# 4. Translate the macro display name
fr = get_macro("MC-10").localized("fr")
print(fr.name)  # "Gouvernance d'Entreprise et Risque"

The model is documented in Section 13 of the governance doc.

Versioning

Two version numbers travel together:

• VERSION — semver of the catalogue content.
• SCHEMA_VERSION — integer; bumps only on non-additive field/value changes.

Pin against SCHEMA_VERSION to detect breaking shape changes; pin against VERSION minor for additive content updates.

Source

The catalogue YAML and build pipeline live at https://github.com/vincentmakes/turbo-ea-capabilities.

License

Dual-licensed, split along the data/code seam:

• The bundled catalogue data (turbo_ea_capabilities/data/*.json) is licensed under CC BY 4.0. Free to use, adapt, and redistribute commercially, provided you credit the source and indicate changes.
• The library code is licensed under MIT.

When you share or adapt the catalogue, credit:

Turbo EA Capabilities by Vincent Verdet — Turbo EA, https://github.com/vincentmakes/turbo-ea-capabilities, CC BY 4.0

The same string is available programmatically — the bundled version.json carries license, license_url, and attribution fields. Full terms: LICENSING.md.