Build a rig

You have a folder where your agent works. Maybe it's an experiment, maybe it's a finished product. Either way, you want to share it — with a teammate, with the world, or with future-you on a different machine.

The progression:

rig init                  # write rig.toml; inspect, ask, infer
rig add data/strategies/  # opt files in beyond the defaults
rig pack                  # produce dist/<name>-<version>.rig.tgz
rig inspect ./dist/...    # sanity check before publishing

After that, you're ready to publish.

Pages in this section

• Initialize (rig init) — the interactive wizard.
• What gets shared by default — the include/exclude model.
• Opt files in (rig add) — when defaults aren't enough.
• Package (rig pack) — produce the artifact.
• Inspect locally — rig inspect your own artifact before publishing.
• Safety: secrets & large files — what the warnings mean.
• Versioning — semver, when to bump, breaking changes.

Related

• Manifest reference — every field of rig.toml.
• Publish — get your rig to other machines.
• Make it collaborative — add a [sync] section.