swarma - Run learning growth experiments with agent teams
Runs agent-team growth experiments, evaluates signals, and builds a validated playbook through repeated cycles.
Tags
Updated: 2026-10-03Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Run growth experiments
- Generate experiment teams
- Test hypotheses at scale
- Evaluate experiment signals
- Compare results with baselines
- Build validated playbooks
- Log external metrics
- Import metric data
- Serve REST or MCP APIs
Inputs
- Growth objective
- Target audience
- Success criteria
- Experiment topic
- External metrics
- Metric CSV files
- OpenRouter API key
Outputs
- Experiment results
- Experiment verdicts
- Team configurations
- Logged metrics
- Evolving strategy files
- Validated playbooks
- Run status
- REST or MCP service
Requirements
- Python 3.11 or later
- pip
- Terminal access
- OpenRouter API key
- LLM access through OpenRouter