Dependency Map
What must happen first before a skill becomes a business?
Critical Path
SPROUT PHASE 1 (prerequisite)
│
▼
PHASE 1: Idea Scoring + Validation
├── 1.1 Wire venture_business_ideas entry point
├── 1.2 AI domain research (eval-runner gate)
├── 1.3 Value proposition generation
├── 1.4 ICP identification (explore-exploit)
├── 1.5 Go/no-go gate (optimal-stopping)
└── 1.6 If pass → trigger Sprout Generator
│
▼
PHASE 2: Operations Seeding
├── 2.1 WorkChart templates per business model
├── 2.2 Sales pipeline seeded
├── 2.3 Content strategy scored (spcl-scoring)
├── 2.4 Accounting structure created
└── 2.5 Discovery session populated
│
▼
PHASE 3: Agent Workforce
├── 3.1 Agent assignment to WorkChart stages
├── 3.2 Human-in-the-loop gates configured
├── 3.3 Agent performance tracking
└── 3.4 AI tuning loop optimization
│
▼
PHASE 4: Network Effects
├── 4.1 Second venture from different domain
├── 4.2 Cross-venture algorithm learning
├── 4.3 Venture-to-venture delegation
└── 4.4 Compound rate tracking
Dependencies by Phase
| Phase | Depends On | Hard/Soft | Unblocks | Effort |
|---|---|---|---|---|
| 0 (Sprout) | Nothing — in progress | — | Everything | Blocked at Step 1.6 |
| 1 (Scoring) | Sprout Phase 1 complete | Hard | Phase 2 + venture creation | 2 weeks |
| 2 (Ops) | Phase 1 + WorkChart engine | Hard | Phase 3 + operational business | 2 weeks |
| 3 (Agents) | Phase 2 + Agent Platform at L2+ | Soft | Phase 4 + automated operations | 3 weeks |
| 4 (Network) | Phase 3 + second venture candidate | Soft | Platform proof | 2 weeks |
Algorithm Dependencies
| Algorithm | Available When | Depends On |
|---|---|---|
| vfl-force-potential | Phase 1.1 | venture_business_ideas wired |
| eval-runner | Phase 1.2 | LLM pipeline established |
| spcl-scoring | Phase 2.3 | Content framework defined |
| explore-exploit | Phase 1.4 | Multiple ICP segments generated |
| optimal-stopping | Phase 1.5 | Research pipeline with measurable signal |
| sales-forecasting | Phase 2.2 | Domain-specific deal data |
| depin-score | Domain-specific | Only for DePIN ventures |
| ai-tuning-loop | Phase 3.4 | Agent performance data accumulated |
| compound-rate-tracking | Phase 4.4 | Multiple ventures with WorkChart receipts |
External Dependencies
| Dependency | Type | Status | Risk |
|---|---|---|---|
| A2A Sprout Generator Phase 1 | Hard prerequisite | Blocked at Step 1.6 | HIGH — nothing starts without this |
| Agent Platform at L2+ | Soft — needed for Phase 3 | L1-L2 (6 new L2 pages) | MEDIUM — progressing independently |
| LLM provider for domain research | Hard — needed for Phase 1 | Available (existing infrastructure) | LOW |
| Real craftsperson for validation | Soft — needed for persona proof | Not started | MEDIUM — no user research planned |
Key Finding
Single-thread critical path through Sprout Phase 1 → Phase 1 scoring pipeline → Phase 2 operations. The bottleneck isn't building the generator — it's validating the prerequisite. Agent Platform and algorithm implementations can progress in parallel, but the end-to-end flow is serial through the first three phases.
Context
- Value Stream Map — Previous: where does time die?
- Capability Map — Next: what can we actually do?
- A2A Sprout Generator — The prerequisite dependency map