Agent Workflow

This is the foundation skill that powers all Maestro commands. It's not user-invocable directly, but it defines the core principles every other command builds on: structured prompts over walls of text, context budgeting instead of dumping entire codebases, focused tool sets instead of exposing 30+ functions. The real value is in the opinionated defaults. It won't let you proceed without understanding your model, task, and priorities first. Includes a "Workflow Slop Test" checklist that catches common mistakes like missing output schemas or multi-agent systems solving single-agent problems. If you're building agent workflows, this is the operating manual baked into the tooling.

skills.sh installs: 374
GitHub stars: 602
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add sharpdeveye/maestro --skill agent-workflow --agent claude-code

Installs into .claude/skills of the current project.

Files

MANDATORY — Context Gathering Protocol

Before applying any workflow guidance, gather context:

Check for Maestro context in the project root
- First check .maestro/context.md (v2 layout)
- Then check .maestro.md (v1 layout — backward compatible)
- If it exists → read it and use the workflow context within
- If it doesn't exist → tell the user: "No workflow context found. Run /teach-maestro to set up project-specific context for better results."
Check for decision history (optional)
- If .maestro/decisions.jsonl exists → read the last 5 decisions for session continuity
- If it doesn't exist → proceed without it (no error)
Minimum viable context (if no .maestro.md):
- What AI model(s) are being used?
- What is the workflow's primary task?
- Are there existing prompts, tools, or agents to work with?
- What are the quality/speed/cost priorities?
DO NOT proceed without at least understanding the model, task, and priorities.