Teach Maestro

What the source does

• FilesystemNone observedNo reads or writes in the source
• NetworkNone observedNo outbound calls in the source
• ShellNone observedNo commands run in the source
• DependenciesPinnedNothing floats
• SecretsNone foundNo keys in source

Five audit layers

Gates applied: no behavioural pass.

Audit 0.4.1 · source sha 4e163a7a8095 · 19 Oct 2026Report an issue or request a re-scan

• Publishersharpdeveye GitHub userOwner of the repository this skill is read from
• Source foldersharpdeveye/maestro/source/skills/teach-maestroThe exact folder we audited
• LicenceMITAs the repository states it
• Commit pinned for this verdict4e163a7a8095A new audit runs when the source changes
• Repository statusNot archivedLast push 173 days ago
• Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

Its description: Use when starting a new project with Maestro or when no .maestro.md context file exists yet. Run once per project.

---
name: teach-maestro
description: "Use when starting a new project with Maestro or when no .maestro.md context file exists yet. Run once per project."
argument-hint: "[project name]"
category: utility
version: 2.0.0
user-invocable: true
---

## BOOTSTRAP — First Command to Run

This is the entry point for Maestro. It creates the `.maestro.md` context file that all other Maestro commands depend on. No other preparation is needed — this IS the preparation.

---

You are conducting a structured interview to understand this project's AI workflow setup. Be conversational but thorough.

### Interview Questions

Ask these questions one section at a time. Wait for answers before proceeding.

**Section 1 — Models & Providers**

- What AI model(s) are you using? (e.g., GPT-4, Claude, Gemini, local models)
- Are you using APIs directly or through a framework? (e.g., LangChain, LlamaIndex, custom)
- What are your context window sizes?

**Section 2 — Workflow Architecture**

- Describe your current workflow at a high level (what goes in, what comes out)
- Do you have multiple agents/steps, or is it a single-agent system?
- What tools/functions are available to your agent(s)?

**Section 3 — Quality & Evaluation**

- How do you currently evaluate output quality?
- Do you have test cases or golden examples?
- What are the most common failure modes?

**Section 4 — Constraints**

- What are your cost constraints? (budget per request, per day)
- What are your latency requirements? (real-time, batch, async)
- Are there compliance requirements? (HIPAA, GDPR, SOC2, etc.)

**Section 5 — Priorities**

- Rank these from most to least important: Quality, Speed, Cost, Safety
- What's the single biggest workflow problem you want to solve?

### Output Format

After gathering all answers, generate a `.maestro.md` file:

```markdown
# Maestro Workflow Context
Generated: [date]

## Models & Providers
[answers from section 1]

## Workflow Architecture
[answers from section 2]

## Quality & Evaluation
[answers from section 3]

## Constraints
[answers from section 4]

## Priorities
[answers from section 5, with ranked priorities]
```

Save this file to the project root as `.maestro.md`.

### Context Completeness

| Section | Status | Impact if Missing |
|---------|--------|-------------------|
| Models & Providers | ? | Commands can't tailor advice to your stack |
| Workflow Architecture | ? | Commands can't assess complexity |
| Quality & Evaluation | ? | `/iterate` and `/evaluate` less effective |
| Constraints | ? | `/guard` and `/accelerate` can't set limits |
| Priorities | ? | All commands default to generic guidance |

### Interview Checklist

- [ ] All 5 sections asked and answered
- [ ] Answers are specific (not vague)
- [ ] `.maestro.md` file generated and saved
- [ ] User confirmed accuracy of the generated context

### Recommended Next Step

After creating `.maestro.md`, run `/diagnose` for a baseline health check of your workflow.

**NEVER**:

- Skip questions — every section matters for downstream commands
- Make assumptions — ask if unclear
- Overwrite an existing `.maestro.md` without asking

Excerpt, first 3,132 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

git clone https://github.com/sharpdeveye/maestro
git -C maestro checkout 4e163a7a8095

Works in

1. 2026-10-19
   ✓SAFE B 89 / 100

   first audit · source 4e163a7a8095

Questions about Teach Maestro

• github.com/sharpdeveye/maestro/tree/main/source/skills/teach-mae — the audited source at commit 4e163a7a8095 and every finding aboveObserved · 19 Oct 2026

Spotted something wrong? Report a correction · Methodology