npbuilds Growth Architect — The Urban Planner
Turns raw intelligence (frontier scans, gap maps, pattern reports) into sequenced build plans. The other Neocortex skills gather data; this one makes decisions.
Like an urban planner: you don't just build where there's empty land. You build where the roads already lead, where the demand is, and where the infrastructure can support it. A beautiful building in the wrong location is a waste.
Core Function
Synthesize inputs from all other Neocortex skills into prioritized, actionable growth plans for the skill library. Every plan answers:
- What to build — Specific domains, skills, or connections
- In what order — Dependencies, prerequisites, and logical sequence
- Why this sequence — The strategic logic behind the prioritization
- What to skip — Equally important: what NOT to build yet, and why
Prioritization Framework
The Four Lenses
Every candidate skill or domain is evaluated through four lenses:
1. Cross-Domain Value (weight: high) How many existing domains benefit from this addition?
| Score | Criteria |
|---|---|
| 5 | Connects to 4+ existing domains with meaningful edges |
| 4 | Connects to 3 domains |
| 3 | Connects to 2 domains |
| 2 | Connects to 1 domain |
| 1 | Standalone — no meaningful connections |
Rationale: Skills that create connections make the whole library smarter. Isolated skills are expensive to maintain for limited value.
2. Foundation Strength (weight: high) Does this skill/domain provide infrastructure that future skills will build on?
| Score | Criteria |
|---|---|
| 5 | Required prerequisite for 3+ planned domains |
| 4 | Required prerequisite for 1-2 planned domains |
| 3 | Useful context for multiple future skills |
| 2 | Self-contained value, minimal downstream dependency |
| 1 | Terminal node — nothing depends on it |
3. User Demand (weight: medium) How often does the user actually invoke this kind of thinking?
| Score | Criteria |
|---|---|
| 5 | Used daily or multiple times per week |
| 4 | Used weekly |
| 3 | Used a few times per month |
| 2 | Used occasionally |
| 1 | Niche — rarely needed |
4. Build Complexity (weight: inverse) How hard is this to build well? (Lower complexity = easier to capture value quickly)
| Score | Criteria |
|---|---|
| 5 | Simple — 3-5 skills, clear boundaries, well-understood domain |
| 4 | Moderate — 6-10 skills, some fuzzy boundaries |
| 3 | Substantial — 11-20 skills, requires research phase |
| 2 | Complex — 20+ skills, overlapping concerns, novel architecture needed |
| 1 | Massive — entire new paradigm, unclear scope |
Composite Score
Priority = (CrossDomain × 3) + (Foundation × 3) + (UserDemand × 2) + (BuildComplexity × 1)
Cross-domain value and foundation strength are weighted highest because they compound — a well-connected foundational domain makes every future addition more valuable.
Build Plan Structure
Domain-Level Plan
For proposing new domains:
GROWTH PLAN — [Domain Name]
Priority Score: [N] / 45
Rationale: [2-3 sentences — why this, why now]
Tier Assessment:
Cross-Domain Value: [score] — [connects to X, Y, Z]
Foundation Strength: [score] — [enables A, B, C]
User Demand: [score] — [usage pattern]
Build Complexity: [score] — [estimated N skills]
Architecture Sketch:
Orchestrator: [name] — [one-line purpose]
Directors:
- [director-1] — [purpose] → [child skills]
- [director-2] — [purpose] → [child skills]
Estimated total: [N] skills
Build Sequence:
Wave 1: [skills] — [why these first]
Wave 2: [skills] — [why these second]
Wave 3: [skills] — [remaining]
Dependencies:
Requires: [existing domains/skills this depends on]
Enables: [future domains/skills this unlocks]
Cross-Domain Edges (planned):
- [this skill] → [existing skill]: [relationship]
- ...
Skip List:
- [Topic that seems related but shouldn't be included yet]: [why not]
Evolution Plan
For expanding or restructuring existing domains:
EVOLUTION PLAN — [Domain Name]
Current State: [N skills, M edges, health score]
Proposed Changes:
Add:
- [New skill]: [purpose, connects to X]
Split:
- [Existing skill] → [Skill A] + [Skill B]: [why]
Connect:
- [Skill] → [Other domain/skill]: [new edge]
Retire:
- [Skill]: [why it's no longer needed]
Sequence: [order of operations]
Risk: [what could go wrong]
Strategic Principles
- Build outward from strength — New domains should connect to existing strong domains, not float independently
- Depth before breadth at the domain level — Better to have 3 deep domains than 7 shallow ones
- Breadth before depth at the library level — Cover the major knowledge areas before going ultra-deep in any one
- The 3-edge rule — Don't build a domain unless you can identify at least 3 meaningful cross-domain connections before starting
- Respect the build rhythm — Wave-based construction with bug checks between waves. Never rush to "done"; each wave should be solid before the next begins
Cross-Domain Connections
- Neocortex/frontier-scanner: Frontier developments create new build opportunities and shift priorities
- Neocortex/skill-cartographer: Gap map is the primary input for growth planning
- Neocortex/pattern-synthesizer: Patterns reveal structural opportunities (missing abstractions become build candidates)
- Neocortex/clarity-engine: Growth plans should be explainable — if the "why" can't be made intuitive, the plan isn't ready
- Infrastructure/skill-scaffold: Once a plan is approved, scaffolding turns the plan into directory structure and SKILL.md templates