Skip to main content

Capability Map

What can we actually do — and where are the gaps?

Capability Maturity

CapabilityMaturityEvidenceCategoryGap?
Venture schema (18 tables × 5 domains)3 — DefinedMigration applied, tables exist in prodCoreNo
WorkChart venture scoping1 — Ad-hoc3 nullable columns added, no data yetCoreP1
Sprout generator3 — DefinedTemplates built, registered, not executedCoreP1
Safe Nx runner4 — ManagedBuilt, tested, handles hang edge caseSupportingNo
Acceptance test script3 — DefinedWritten, 7-line spec, not executedSupportingNo
Venture self-creation (boot)2 — RepeatableDesigned, coded in template, untestedCoreP1
Agent card auto-registration2 — RepeatableDesigned, coded in template, untestedCoreP1
A2A endpoints1 — Ad-hocStubs onlyCoreP2
Feature scaffold (per-entity CRUD)3 — DefinedWorks for drmg-sales, untested standaloneSupportingP2
Shared lib independence0 — Not BuiltNever tested outside drmg-salesCoreP1
Domain-specific entities0 — Not BuiltNo tables for spots/reports/catchesGenericP2
Venture access control0 — Not BuiltNo design yetSupportingP3
Cross-venture A2A0 — Not BuiltRequires two working venturesCoreP3

Maturity Distribution

LevelCount%
0 — Not Built323%
1 — Ad-hoc215%
2 — Repeatable215%
3 — Defined431%
4 — Managed18%
5 — Optimised00%

Average maturity: 1.8 — between Ad-hoc and Repeatable. The generator lifts the average significantly once run.

Investment Strategy

CategoryCapabilitiesStrategy
Core (differentiating)Venture schema, WorkChart scoping, generator, self-creation, agent registration, shared lib independence, cross-venture A2AInvest — these ARE the platform thesis
Supporting (necessary)Safe Nx runner, acceptance test, feature scaffold, access controlMaintain — solve once, reuse everywhere
Generic (commodity)Domain-specific entitiesOutsource to generator templates

Key Finding

The riskiest capability is also the most revealing: shared lib independence (maturity 0). Everything else could score 5 and the platform thesis still fails if shared libs carry hidden drmg-sales coupling. The generator run (Phase 1 Step 1.6) is simultaneously the most important test and the most dangerous one.

Context

  • Outcome Map — What success looks like when these gaps close
  • Dependency Map — Which gaps block which phases
  • A&ID — How agents and instruments orchestrate these capabilities