Skip to content

Latest commit

Β 

History

History
74 lines (51 loc) Β· 2.35 KB

File metadata and controls

74 lines (51 loc) Β· 2.35 KB

Learning Path

This guide gives Swarm Command readers a sane order to follow instead of dropping straight into deep protocol docs.


Path 1 β€” I just want to try it (5 minutes)

  1. Read the 30-Second Overview in the README
  2. Copy the install command from the README
  3. Run the Quick Demo prompt
  4. Skim use-cases.md for your next prompt

Best for: first-time users, evaluators, demos.


Path 2 β€” I want to use it well (15 minutes)

  1. Read the README
  2. Read scaling.md
  3. Read use-cases.md
  4. Skim architecture.md
  5. Return to the README FAQ when questions come up

Best for: operators, engineers deciding between SS-50 / SS-100 / SS-250.


Path 3 β€” I want to understand the system design (25 minutes)

  1. Read the architecture summary in the README
  2. Read architecture.md
  3. Open architecture-diagrams.md
  4. Read consensus.md
  5. Read shadow-scoring.md
  6. Finish with scaling.md

Best for: architects, contributors, people adapting the pattern to another swarm system.


Path 4 β€” I want the origin story

  1. Read How It Was Built in the README
  2. Read the motivation section in shadow-scoring.md
  3. Read architecture.md with the failure modes in mind

Best for: people who care how the system evolved, not just what it does.


Suggested reading order by persona

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/

One-sentence summaries for each doc

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