Open-source AI operating system for validating ideas and executing goals. 3 agents, 15 skills.
Define goals, research markets, reason through problems — all from the command line.
| Agent | What It Does |
|---|---|
fnd-architect |
Sets up strategic foundation — business mode, constraints, context |
fnd-researcher |
Market research — TAM/SAM/SOM, segments, problem validation, competitors |
rsn-problem-solver |
Structured reasoning — causal, abductive, inductive, analogical, dialectical, counterfactual |
System backbone (6) — goal-driven execution:
| Skill | What It Does |
|---|---|
sys-defining-goals |
Transform intent into measurable goal files |
sys-decomposing-goals |
Break goals into subgoal hierarchies |
sys-activating-goals |
Create execution threads from goals |
sys-tracking-goals |
Monitor progress against targets |
sys-executing-threads |
Write agent outputs to thread stage files |
sys-indexing-directories |
Generate index files for context loading |
Reasoning (4) — structured thinking:
| Skill | What It Does |
|---|---|
rsn-reasoning-problems |
6 cognitive modes for problem solving |
rsn-perceiving-information |
Systematic information gathering |
rsn-creating-ideas |
Creative ideation and novel solutions |
rsn-learning-outcomes |
Extract insights from outcomes |
Foundations research (5) — idea validation:
| Skill | What It Does |
|---|---|
fnd.r-sizing-markets |
Calculate TAM/SAM/SOM |
fnd.r-segmenting-customers |
Define customer segments and ICP |
fnd.r-scoring-problems |
Score problem severity and willingness to pay |
fnd.r-analyzing-competition |
Map competitive landscape |
fnd-validating-gates |
Validate phase transition requirements |
git clone https://github.com/BellaBe/lean-os.git my-project
cd my-projectrun fnd-architect agent
This creates your business mode, captures constraints, and initializes strategy/canvas/.
run fnd-researcher agent
Sizes your market, defines segments, scores the problem, maps competitors.
run skill sys-defining-goals
Creates a measurable goal in strategy/goals/active/.
Goals create threads. Threads follow a 6-stage causal flow:
1-input → 2-hypothesis → 3-implication → 4-decision → 5-actions → 6-learning
Every action links to a goal. Every outcome captures learning.
You (decisions)
|
Agents (orchestration)
|
Skills (execution)
Goals drive threads. Threads follow a 6-stage causal flow. Agents execute threads by calling skills. The system reads index files on start for context.
my-project/
├── strategy/
│ ├── canvas/ # Business canvas
│ └── goals/ # Goal definitions
├── threads/ # Execution threads (6-stage causal flow)
├── artifacts/ # Deliverables
├── docs/ # Documentation
└── .claude/
├── agents/ # 3 orchestrators
└── skills/ # 15 skills
Set in strategy/canvas/00.mode.md:
| Mode | Optimizes For |
|---|---|
| BOOTSTRAP | Profitability, cash flow, fast decisions |
| VENTURE | Growth rate, market size, defensibility |
Core validates your idea. Pro builds and runs the business.
| Capability | Core | Pro |
|---|---|---|
| Agents | 3 | 39 |
| Skills | 15 | 191 |
| Goal system | Full | Full |
| Reasoning | Full (6 modes) | Full + domain intelligence |
| Market research | Full | Full |
| Business modeling | -- | Pricing, unit economics, cost structure |
| GTM planning | -- | Channels, assumptions, launch plan |
| Engineering | -- | Backend IR pipeline, frontend, Shopify |
| Design systems | -- | Intent to web/infographic specs (16 skills) |
| Sales | -- | Outbound, partnerships, enablement |
| Marketing | -- | Inbound, content, campaigns |
| Customer success | -- | Onboarding, retention, expansion |
| RevOps | -- | Signals, scoring, allocation |
| Product | -- | Features, PLG, growth mechanics |
| Critique | -- | Quality validation across artifacts |
| Behavioral science | -- | Choice architecture, nudges |
MIT. See LICENSE.