A Lemon Squeezy CLI for developers and AI agents

UX Canvas runs on Lemon Squeezy for billing, and I kept hitting the same friction: needing to log in to the dashboard to grab an ID, check a subscription status, or trigger some side effect. Every time, it pulled me out of flow. I wanted my AI agents — which already handle most of my development workflow — to manage the billing platform too, without me in the loop.

lmsq is a complete CLI for the Lemon Squeezy API — 62 operations across 22 resource types covering stores, orders, subscriptions, license keys, checkouts, webhooks, and everything else the API exposes. It auto-detects whether it's being called by a human in a terminal or piped to an AI agent, and adjusts its output format accordingly — compact key-value pairs that respect context windows when piped, colored tables when interactive. One npx skills add command installs an agent skill that teaches any AI assistant (Cursor, Claude Code, OpenCode, Codex) how to use every command. I built it with Opus 4.6 using TDD and verification loops to keep quality high across all 59 tests.