What problem does it solve? Writing idiomatic re-frame2 ClojureScript code requires knowing the framework's canonical declarations, decision rules, and 14 canonical patterns; this Skill supplies those recipes so an AI agent produces correct events, subscriptions, effects, frames, state machines, schemas, stories, and routing on the first attempt. ## Core Features & Use Cases - Canonical pattern recipes: Provides one leaf per canonical pattern (RemoteData, Resources, Forms, Boot, WebSocket, NineStates, ManagedHTTP, AsyncEffect, StaleDetection, and more) with load triggers, mini-declarations, and worked-example links. - Decision trees: Routes authoring tasks through pick-a-pattern and slice-or-machine decision trees to choose between slices, regions, and state machines. - Reference depth: Carries fundamentals, state-machine, tooling, and cross-cutting references (testing, privacy/elision, SSR authoring, observability) loaded on demand, at most two leaves per task. - Use Case: Ask the agent to write a reg-machine for a login lifecycle with a parallel region for password-visibility toggling, and it loads the machine grammar and Forms pattern leaves, matches the canonical declaration shape, and runs the project's declared test gate before reporting done. ## Quick Start Ask the agent to write a re-frame2 event handler, subscription, or state machine for your feature, for example by mentioning reg-event, reg-sub, dispatch, or a pattern name like RemoteData in your prompt.