@DanKornas: AI coding agents don’t need more random prompts. They need a workflow layer. Maestro is an open-source workflow system …

AI coding agents don’t need more random prompts. They need a workflow layer. Maestro is an open-source workflow system for AI coding agents, packaged as skills, commands, reference files, memory, and an MCP server for AI coding tools. It helps you make agent work more consistent by giving the model project context, reusable commands, domain references, persistent session records, and explicit anti-patterns instead of starting every session from scratch. Key features: • 25 workflow commands – diagnose, evaluate, refine, streamline, fortify, capture, reflect, and more • Agent-workflow skill – one core skill backed by 7 references for prompts, context, tools, architecture, feedback, knowledge systems, and guardrails • Persistent memory layer – decisions, audit trail, and session history can survive across sessions • Context gathering protocol – uses .maestro.md or .maestro/context.md so commands have project-specific awareness • Multi-tool support – works across Cursor, Claude Code, Gemini CLI, Codex CLI, Copilot, Kiro, Trae, OpenCode, Pi, and more It’s open-source (MIT license). Link in the reply