Foundations Workflow
How to set up and validate your business canvas using the Foundations agents and skills.
Workflow Selection
START
|
+--> "I have a new idea"
| +--> Quick Validation workflow
|
+--> "I want to build a canvas"
| +--> Step by step: fnd-architect --> fnd-researcher
|
+--> "I need to work on a specific section"
+--> See Single Block Reference below
Phase Execution
Two phases available in Core. Each has a dedicated agent.
fnd-architect fnd-researcher
Phase 0: Setup Phase 1: Discovery
00, 01, 02 03, 04, 05, 06
| |
G0 G1
|
v
Goal Initialization
(sys-defining-goals)
Phase 0: Setup (fnd-architect)
Process:
- Check existing state in
strategy/canvas/ - Interview for mode: growth priority, funding intent, risk tolerance
- Interview for KBOS: Known facts, Beliefs, Observations, Strategic intent
- Interview for constraints: budget, timeline, resources, technical, regulatory
- Validate G0
Mode determines downstream thresholds:
| Mode | TAM Requirement | LTV:CAC Min | Payback Max |
|---|---|---|---|
| VENTURE | > $1B | 2:1 (early OK) | 24 mo |
| BOOTSTRAP | Any (SOM focus) | 3:1 (now) | 12 mo |
Phase 1: Discovery (fnd-researcher)
| Step | Skill | Output | Quality Check |
|---|---|---|---|
| 1 | fnd.r-sizing-markets | 03.opportunity.md | TAM has cited source, SOM realistic |
| 2 | fnd.r-segmenting-customers | 04.segments.md | Filters are searchable, pain scores have evidence |
| 3 | fnd.r-scoring-problems | 05.problem.md | Top 3 with FIWTP, evidence cited |
| 4 | fnd.r-analyzing-competition | 06.competitive.md | 3+ competitors, positioning matrix, gaps |
Gate validation: fnd-validating-gates checks phase transition requirements.
After G1: What's Next
After Discovery, you have a validated market understanding. From here:
- Define goals from what you've learned:
sys-defining-goals - Upgrade to Pro for business modeling (pricing, unit economics), GTM planning, and full 16-section canvas
Quick Validation
When: New idea, considering a pivot, before writing code, testing multiple ideas.
Outputs: Minimal canvas (04, 05) + validation direction.
Process:
- Define segment — specific, findable customer with 1-2 observable filters
- Validate problem — evidence from forums, job postings, competitor reviews
- Design validation approach — test riskiest assumption first
Decision after validation:
Evidence supports the problem?
+--> YES: Proceed to full Discovery (fnd-researcher)
+--> MIXED: Refine segment or problem, gather more evidence
+--> NO: Pivot segment, pivot problem, or kill idea
Single Block Reference
| Block | File | Skill/Agent | Prompt |
|---|---|---|---|
| Mode | 00.mode | fnd-architect | "Help me define my business mode" |
| Context | 01.context | fnd-architect | "Document my strategic context" |
| Constraints | 02.constraints | fnd-architect | "Define my constraints" |
| Opportunity | 03.opportunity | fnd.r-sizing-markets | "Size my market" |
| Segments | 04.segments | fnd.r-segmenting-customers | "Define customer segments" |
| Problems | 05.problem | fnd.r-scoring-problems | "Validate problems" |
| Competitive | 06.competitive | fnd.r-analyzing-competition | "Analyze competition" |
Sections 07-15 (value proposition, pricing, unit economics, GTM) require LeanOS Pro.
Post-Discovery: Goal Initialization
After Discovery (G1 passed), initialize goals from what you've learned.
Discovery Complete (G1)
|
v
READ canvas sections
- 00.mode.md --> metric framework
- 04.segments.md --> target customer
- 05.problem.md --> validated problems
|
v
INVOKE sys-defining-goals
- Define your first measurable goal
|
v
INVOKE sys-decomposing-goals (optional)
- Break into subgoals
|
v
INVOKE sys-activating-goals
- Create execution thread
- Begin causal flow
Thread Structure
All Foundations workflows use causal flow threads:
threads/foundations/{workflow}/{context}/
+-- 1-input.md # Canvas state, user request, constraints
+-- 2-hypothesis.md # Approach, assumptions about market/customer
+-- 3-implication.md # If hypothesis true, what canvas sections follow
+-- 4-decision.md # Chosen direction, rationale
+-- 5-actions.md # Phase execution, skill invocations
+-- 6-learning.md # Validation results, pivots, insights