Add the NomadWorks plugin package to the environment where OpenCode resolves plugins.
{
"plugin": ["@neuralnomads/nomadworks"]
}OpenCode's config schema allows plugin entries to be strings or [package, options] tuples. Use the form that matches your broader OpenCode setup.
Add NomadWorks to your OpenCode configuration file and restart OpenCode so the plugin is loaded.
Minimal example:
{
"$schema": "https://opencode.ai/config.json",
"plugin": ["@neuralnomads/nomadworks"]
}After the plugin loads, open the repository you want to enable and start talking to the product_manager agent (PMA).
PMA is the default orchestrator and will help you set up the repository for NomadWorks. If the repository has not been initialized yet, PMA can trigger the required setup flow on your behalf.
During setup, PMA should ask whether you want:
miniteam mode for simple repositories andtiny/standardworkfullteam mode for the complete collective, includingcomplexworkflows
You do not need to manually run NomadWorks initialization commands as a first step.
When PMA initializes the repository, NomadWorks creates:
.nomadworks/nomadworks.yaml.nomadworks/policies/README.md.nomadworks/agents/README.md.nomadworks/agent-additions/README.md.nomadworks/generated/agents/README.md.nomadworks/generated/policies/README.mdcodemap.ymltasks/current.mdtasks/done.mddocs/scrs/current.mddocs/scrs/done.md
Edit .nomadworks/nomadworks.yaml to set defaults, features, policy extraction behavior, and per-agent config overrides. Use .nomadworks/agents/ for full repository-local agent definitions or custom agents, and .nomadworks/agent-additions/ for additive repo-specific agent instructions.
See:
docs/setup/CONFIGURATION.mddocs/guides/AGENTS.mddocs/guides/WORKFLOW.md