Living Brain Service CLI

Command-line interface for the Living Brain Service API.

Table of contents

• Installation
• Authentication
• Quick start
• Usage
• Documentation
• Advanced
  • Common flags
  • Environment variables
  • Output formats
  • Shell completion
• Attribution

Installation

Shell (macOS / Linux)

curl --proto '=https' --tlsv1.2 -LsSf https://github.com/ChatAndBuild/livingbrain-cli/releases/latest/download/livingbrain-cli-installer.sh | sh

PowerShell (Windows)

powershell -ExecutionPolicy ByPass -c "irm https://github.com/ChatAndBuild/livingbrain-cli/releases/latest/download/livingbrain-cli-installer.ps1 | iex"

Build from source

If you prefer to build from source, install the Rust toolchain and run:

cargo build --release
./target/release/livingbrain --help

Authentication

Every request needs two values. Set them as environment variables, or pass the matching flag on any command:

Value	Env var	Flag	Header sent
API key	LIVING_BRAIN_API_KEY	--api-key	x-api-key
Subject ID (required)	LIVING_BRAIN_SUBJECT_ID	--subject-id	x-subject-id

The subject ID identifies the end user you're acting on behalf of and must be a UUID. Brains are owned by a subject, so it determines which brains the CLI can see.

export LIVING_BRAIN_API_KEY="<your api key>"
export LIVING_BRAIN_SUBJECT_ID="3f2b8c1e-7a4d-4e9b-9c2a-5d6e7f8a9b0c"

A .env file in the working directory is also supported — the CLI auto-loads it on startup.

Quick start

List available commands:

livingbrain --help

Call an API endpoint:

livingbrain <resource> <method>

Run livingbrain <resource> --help to see available methods for a resource.

Usage

Every API resource appears as a subcommand (e.g. livingbrain <resource> <method>). Run livingbrain <resource> --help to see available methods.

Provide request parameters as flags or as JSON:

livingbrain <resource> <method> --json '{"key": "value"}'

Documentation

See reference.md for the full command reference.

Advanced

Common flags

These flags are available on every operation:

Flag	Description
--dry-run	Validate the request locally and print the HTTP request without sending it
--json <JSON|->	Supply a request body as JSON (or - to read stdin)
--params <JSON>	Merge extra parameters as JSON (overrides individual flags)
--format <json|table|yaml|csv>	Output format (default json)
--output <PATH>	Write binary responses to a file
--base-url <URL>	Override the API base URL
--no-extract	Print the full response body instead of the x-fern-sdk-return-value extraction
--no-retry	Disable retries declared by x-fern-retries, including network errors
-q, --quiet	Suppress stdout output on success (errors still go to stderr)

Operations the spec describes how to page (via x-fern-pagination or a root page_token parameter) also accept:

Flag	Description
--page-all	Auto-paginate and stream results as NDJSON
--page-limit <N>	Max pages to fetch when auto-paginating (default 10)
--page-delay <MS>	Delay between page fetches in milliseconds (default 100)
--no-pager	Disable the pager even on interactive terminals

Operations the spec marks as streaming (via x-fern-streaming) also accept:

Flag	Description
--no-stream	Buffer the streaming response and print it as a single value once complete

Environment variables

Variable	Description
LIVINGBRAIN_BASE_URL	Override the API base URL
LIVINGBRAIN_CA_BUNDLE	Path to PEM file with extra trust roots (or SSL_CERT_FILE)
LIVINGBRAIN_INSECURE=1	Skip TLS verification (debugging only)
LIVINGBRAIN_PROXY	HTTP(S) proxy URL
LIVINGBRAIN_TIMEOUT_SECS	Total request timeout in seconds

Standard environment variables (HTTPS_PROXY / HTTP_PROXY / NO_PROXY / SSL_CERT_FILE) are also honored.

Output formats

Use the global --format flag to control output. Supported values: json, table, yaml, csv, jsonl, raw, http.

Without --format, output (including errors) is table when stdout is a terminal and json when it is piped or redirected — so scripts and agents get JSON by default. Pass --human to keep the interactive rendering when piping to a pager, and --format json to pin JSON in a terminal.

# Pipe JSON output through jq
livingbrain <resource> <method> --format json | jq

# Keep the human rendering even when piped
livingbrain <resource> <method> --human | less

# Machine-readable catalog of every operation (same as --schema)
livingbrain --help --format json | jq '.operations | length'

Shell completion

Generate shell completion scripts:

livingbrain completion <bash|zsh|fish|powershell>

Attribution

Built on fern-cli-sdk, Copyright Fern, licensed under the Apache License 2.0.