This guide gives Swarm Command readers a sane order to follow instead of dropping straight into deep protocol docs.
- Read the 30-Second Overview in the README
- Copy the install command from the README
- Run the Quick Demo prompt
- Skim use-cases.md for your next prompt
Best for: first-time users, evaluators, demos.
- Read the README
- Read scaling.md
- Read use-cases.md
- Skim architecture.md
- Return to the README FAQ when questions come up
Best for: operators, engineers deciding between SS-50 / SS-100 / SS-250.
- Read the architecture summary in the README
- Read architecture.md
- Open architecture-diagrams.md
- Read consensus.md
- Read shadow-scoring.md
- Finish with scaling.md
Best for: architects, contributors, people adapting the pattern to another swarm system.
- Read How It Was Built in the README
- Read the motivation section in shadow-scoring.md
- Read architecture.md with the failure modes in mind
Best for: people who care how the system evolved, not just what it does.
| Persona | Start | Next | Finish |
|---|---|---|---|
| New user | README | use-cases.md | scaling.md |
| Operator | README | scaling.md | FAQ + architecture.md |
| Architect | architecture.md | consensus.md | shadow-scoring.md |
| Contributor | README | architecture.md | templates/ + protocols/ |
| Doc | Why it exists |
|---|---|
README.md |
The best single-page overview |
architecture.md |
The system model in prose |
architecture-diagrams.md |
The visual version of the architecture |
scaling.md |
How to choose a scale and what it costs |
use-cases.md |
Prompts that make the swarm feel concrete |
consensus.md |
How outputs are merged and scored |
shadow-scoring.md |
How hidden validation catches false confidence |