Guides

Practical guides for using SWARM effectively.

• :material-file-cog: Writing Scenarios

  ---

  Define custom scenarios with YAML configuration

• :material-robot: Custom Agents

  ---

  Implement your own agent behaviors

• :material-chart-scatter-plot: Parameter Sweeps

  ---

  Run systematic experiments across parameter spaces

• :material-brain: LLM Agents

  ---

  Integrate large language models as agents

• :material-shield-alert: Red Teaming

  ---

  Adversarial testing and attack scenarios

• :material-flask: Research Workflow

  ---

  Multi-agent workflow for rigorous SWARM research

• :material-tune: Custom Governance Levers

  ---

  Configure and combine governance mechanisms to control ecosystem health

• :material-transfer: Transferability Considerations

  ---

  Understand when and how simulation results generalize to real systems

Guide Overview

Guide	Description	Difficulty
Writing Scenarios	YAML-based scenario definition	Beginner
Custom Agents	Extend BaseAgent for custom behavior	Intermediate
Parameter Sweeps	Systematic experimentation	Intermediate
LLM Agents	GPT/Claude integration	Advanced
Red Teaming	Adversarial evaluation	Advanced
Research Workflow	Multi-agent research with depth/breadth control	Advanced
Custom Governance Levers	Configure governance mechanisms	Advanced
Transferability Considerations	When results generalize	Advanced