👋 Welcome to Maestro — Workflow Fluency for AI Agents

[sharpdeveye]

👋 Welcome to Maestro

Hey everyone — thanks for checking out Maestro! This is the official discussion space for the project. Whether you're building AI agents, optimizing LLM workflows, or just exploring — you're in the right place.

What is Maestro?

Maestro is an open-source AI workflow skill — a structured system of 21 commands and curated anti-patterns that turns chaotic AI agent development into a repeatable, production-grade process.

npx skills add sharpdeveye/maestro

One command. 21 workflow commands. 7 reference domains. Zero configuration.

---

🧩 What You Get

Category	Commands	What It Does
Build	7 commands	Design prompts, tools, agents, and knowledge systems from scratch
Analyze	4 commands	Audit workflow quality with scored dimensions
Optimize	4 commands	Refine prompts, context windows, and tool chains
Fix	5 commands	Harden with guardrails, retries, and zero-defect verification
Ship	1 command	Deploy with confidence

🆕 Latest: /zero-defect (v1.3.1)

Our newest command — a model-agnostic precision protocol that enforces 8 execution rules, pre-commit verification gates, and anti-pattern correction tables. Think of it as a quality gate that works with any AI model.

---

🗣️ Discussion Topics

We'd love to hear from you. Here are some threads worth starting:

• 🛠️ Show Your Workflow — Share how you're using Maestro in your projects
• 💡 Feature Requests — What commands or domains would you add?
• 🐛 Bug Reports — Something broken? Let us know
• 📖 Use Cases — What AI agent challenge brought you here?
• 🤝 Integrations — Using Maestro with Cursor, Windsurf, Cline, Claude Code, or Antigravity? Share your setup

---

🔗 Quick Links

Resource	Link
🌐 Live Showcase	maestroskills.dev
📦 npm Package	npmjs.com/package/maestro
📖 README	GitHub README
🐛 Issues	Report a Bug

---

🚀 Getting Started in 30 Seconds

# 1. Install
npx skills add sharpdeveye/maestro

# 2. Initialize your project context
/teach-maestro

# 3. Audit your workflow
/diagnose

# 4. Fix and harden
/fortify
/zero-defect

# 5. Ship it
/ship

---

⭐ Support the Project

If Maestro helps your workflow, consider:

• ⭐ Starring the repo — it helps others discover the project
• 💬 Starting a discussion — share what you're building
• 🍴 Contributing — PRs welcome for new anti-patterns, commands, or docs

---

🏷️ Labels for Discussion Categories

We recommend setting up these categories:

Category	Description
📣 Announcements	Updates, releases, and news
💡 Ideas	Feature requests and suggestions
🙏 Q&A	Questions and help
🛠️ Show & Tell	Share your Maestro-powered workflows
📚 Resources	Tutorials, guides, and tips

---

Looking forward to building with you all 🙌

---

Maestro — 1 skill, 21 commands, zero workflow slop.

[nonfuntoke]

Appreciate the discussion here — the agent angle stood out as especially practical.

A small pattern that may help:

• separate discovery, routing, and publish state so the workflow can adapt to issue/comment/discussion surfaces
• keep a compact evidence block for why a surface was selected

If useful, we keep a structured index of similar workflows here: https://skillslookup.com

Happy to share more detail if it would help.

[sharpdeveye]

Appreciate the insights! The 'compact evidence block' is actually spot on with the v2 memory layer I'm architecting right now—forcing the agent to log the why in a decision ledger before executing.

I'll definitely check out the directory. How are you currently handling the routing state on your end? Are you using standard state machines or something more dynamic?