Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when starting a new project with Maestro or when no .maestro.md context file exists yet. Run once per project.
.claude/skills/sharpdeveye-teach-maestro/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 5% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 2% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -21% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -25% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -32% | 0% |
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.
Ask these questions one section at a time. Wait for answers before proceeding.
Section 1 — Models & Providers
Section 2 — Workflow Architecture
Section 3 — Quality & Evaluation
Section 4 — Constraints
Section 5 — Priorities
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.
| 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 |
.maestro.md file generated and savedAfter creating .maestro.md, run /diagnose for a baseline health check of your workflow.
NEVER:
.maestro.md without asking| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | pass→fail | 19,255 | 11,660 | -39% | 1 | 1 | 0% | 3,581 | 2,640 | -26% | 0 | 0 | — |
case-01 | fail→pass | 7,031 | 3,458 | -51% | 1 | 1 | 0% | 1,066 | 1,119 | +5% | 0 | 0 | — |
case-02 | fail→pass | 7,101 | 2,982 | -58% | 1 | 1 | 0% | 1,098 | 1,122 | +2% | 0 | 0 | — |
case-22 | pass→fail | 17,568 | 10,502 | -40% | 1 | 1 | 0% | 3,079 | 2,225 | -28% | 0 | 0 | — |
case-03 | fail→pass | 8,088 | 3,069 | -62% | 1 | 1 | 0% | 1,338 | 1,053 | -21% | 0 | 0 | — |
case-04 | fail→fail | 5,230 | 3,153 | -40% | 1 | 1 | 0% | 855 | 1,166 | +36% | 0 | 0 | — |
case-05 | fail→pass | 8,866 | 3,090 | -65% | 1 | 1 | 0% | 1,534 | 1,150 | -25% | 0 | 0 | — |
case-06 | fail→pass | 8,860 | 2,605 | -71% | 1 | 1 | 0% | 1,495 | 1,023 | -32% | 0 | 0 | — |
case-07 | pass→fail | 9,200 | 5,319 | -42% | 1 | 1 | 0% | 1,537 | 1,455 | -5% | 0 | 0 | — |
case-08 | pass→pass | 6,802 | 3,059 | -55% | 1 | 1 | 0% | 1,036 | 1,097 | +6% | 0 | 0 | — |
case-09 | fail→pass | 8,043 | 3,241 | -60% | 1 | 1 | 0% | 1,215 | 1,108 | -9% | 0 | 0 | — |
case-10 | fail→fail | 13,934 | 3,700 | -73% | 1 | 1 | 0% | 2,345 | 1,362 | -42% | 0 | 0 | — |
case-11 | fail→pass | 9,199 | 2,469 | -73% | 1 | 1 | 0% | 1,451 | 1,073 | -26% | 0 | 0 | — |
case-12 | fail→pass | 7,538 | 2,431 | -68% | 1 | 1 | 0% | 1,215 | 1,031 | -15% | 0 | 0 | — |
case-13 | fail→pass | 12,077 | 7,475 | -38% | 1 | 1 | 0% | 1,663 | 1,760 | +6% | 0 | 0 | — |
case-14 | fail→pass | 3,531 | 3,181 | -10% | 1 | 1 | 0% | 528 | 1,153 | +118% | 0 | 0 | — |
case-15 | fail→pass | 8,681 | 4,053 | -53% | 1 | 1 | 0% | 1,255 | 1,276 | +2% | 0 | 0 | — |
case-16 | fail→pass | 2,925 | 4,502 | +54% | 1 | 1 | 0% | 377 | 1,357 | +260% | 0 | 0 | — |
case-17 | fail→pass | 10,338 | 3,200 | -69% | 1 | 1 | 0% | 1,552 | 1,116 | -28% | 0 | 0 | — |
case-18 | fail→pass | 6,468 | 4,141 | -36% | 1 | 1 | 0% | 1,011 | 1,330 | +32% | 0 | 0 | — |
case-19 | fail→pass | 6,696 | 2,794 | -58% | 1 | 1 | 0% | 973 | 1,054 | +8% | 0 | 0 | — |
case-20 | pass→pass | 13,924 | 6,831 | -51% | 1 | 1 | 0% | 2,181 | 1,741 | -20% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +55 percentage points is the difference between those two pass rates over the 22 comparable cases. 3 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.