CLI

The rngo CLI manages project configurations, runs simulations and routes effect streams to systems.

The source code can be found here: https://github.com/rngodev/cli

Installation

You can install the rngo CLI on macOS and Linux using Homebrew:

brew install rngodev/tap/cli

Otherwise, you can pull down a precompiled binary from https://github.com/rngodev/cli/releases/latest

Additionally, Rust users can build from source via:

cargo install rngo-cli

Most commands interact with the API, so you'll need to run the login command before doing anything else.

The configuration for a project lives in its local .rngo directory.

The main configuration file is .rngo/config.yml, which may look something like this:

key: blog
seed: 41
start: now - years(3)
ai:
  agent: claude

Key		Description
`key`	required	The unique human-readable key of the simulation. rngo sim init sets this to the project's directory name. Must be unique within your rngo account.
`seed`	optional	The default seed for the simulation's RNG, ensuring reproducible runs.
`start`	optional	The starting point in time for the simulation. Accepts expressions like`now - weeks(8)`.
`ai.agent`	optional	The AI agent used for inference tasks. Supported values include `claude`, `codex` or `copilot`.

Systems and effects each get their own configuration files under .rngo/systems/ and .rngo/effects/, e.g.:

> ls -l .rngo
.rngo/config.yml
.rngo/systems/db.yml
.rngo/effects/user.create.yml
.rngo/effects/post.create.yml
.rngo/effects/comment.create.yml

These may be inferred by rngo system infer and rngo effect infer, but see the System and Effect references for how to customize.