EliaAlberti

Scaffold a new Deep Agents (LangChain) agent, a dcode CLI agent, or both, from one command. Use ONLY when the user explicitly runs /new-dcode-agent; never auto-trigger. It interviews the user (form, name, purpose, tools, model, safety), shows a spec, and on confirmation writes a self-contained agent into the user's current project (and/or a dcode CLI agent under ~/.deepagents). The agents it writes work with any OpenAI-compatible API via environment variables.

Run a heuristic UX audit of a customer journey from screenshots. Use when the user asks for a UX audit, heuristic evaluation, usability review, design teardown, or expert review of screens, mockups, or a flow, whether your own product or a competitor's. Evaluates against named frameworks (Nielsen, Shneiderman, Gerhardt-Powals, Bastien & Scapin, Hick/Fitts/Miller/Jakob/Peak-End, Fogg, Cialdini, Gestalt, Norman, Tognazzini, WCAG 2.1, content design heuristics) and produces severity-rated findings with heuristic citations, annotated screenshots, and a structured markdown report. Do not use for generating new designs or for code-level accessibility scans of a live DOM.