Install any skill in seconds. Free to start, no credit card required.
Get Started Free →The single entry point to the Conductor system - state your goal and everything is handled automatically
.claude/skills/ibrahim-3d-go/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | -54% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -33% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -66% | 0% |
The single entry point to the entire Conductor system.
Just state your goal. The system handles everything else.
/go <your goal>/go Add Stripe payment integration
/go Fix the login bug where users get logged out
/go Build a dashboard with analytics
/go Refactor the asset generation to use cachingWhen you invoke /go, follow this process:
Parse the user's goal from $ARGUMENTS:
Check conductor/tracks.md for matching existing tracks:
conductor/tracks/{goal-slug}_{date}/spec.md from the goalplan.md with DAGmetadata.json with v3 schema AND set superpower_enhanced: true (new tracks use superpowers by default)Example metadata.json:
json{ "version": 3, "track_id": "goal-slug_20260213", "type": "feature", "status": "new", "superpower_enhanced": true, "loop_state": { "current_step": "NOT_STARTED", "step_status": "NOT_STARTED" } }
Invoke the conductor-orchestrator agent:
Use the conductor-orchestrator agent to run the evaluate-loop for this track.The orchestrator will:
superpower_enhanced flag to determine which agents to use:Behavior depends on conductor/config.json → "mode":
"agentic" (default) — Fully autonomous. Never stops to ask the user. All decisions resolved by agents, leads, and board."human-in-the-loop" — Pauses at decision points (ambiguity, blockers, fix limits, high-impact decisions) to ask the user.See conductor/config.json for current mode. Edit to switch.
/go # Continues the active track
/go continue # Same as above| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 21,802 | 6,781 | -69% | 1 | 1 | 0% | 4,364 | 936 | -79% | 0 | 0 | — |
case-02 | fail→fail | 8,536 | 6,639 | -22% | 1 | 1 | 0% | 1,243 | 1,156 | -7% | 0 | 0 | — |
case-03 | fail→fail | 3,262 | 3,939 | +21% | 1 | 1 | 0% | 204 | 887 | +335% | 0 | 0 | — |
case-04 | fail→fail | 4,176 | 4,994 | +20% | 1 | 1 | 0% | 546 | 842 | +54% | 0 | 0 | — |
case-05 | fail→pass | 14,408 | 2,126 | -85% | 1 | 1 | 0% | 2,251 | 1,030 | -54% | 0 | 0 | — |
case-06 | fail→pass | 10,929 | 2,725 | -75% | 1 | 1 | 0% | 1,758 | 1,181 | -33% | 0 | 0 | — |
case-07 | fail→pass | 7,005 | 2,699 | -61% | 1 | 1 | 0% | 987 | 1,096 | +11% | 0 | 0 | — |
case-08 | fail→fail | 9,491 | 1,992 | -79% | 1 | 1 | 0% | 1,297 | 1,033 | -20% | 0 | 0 | — |
case-09 | pass→pass | 11,241 | 5,817 | -48% | 1 | 1 | 0% | 1,528 | 1,558 | +2% | 0 | 0 | — |
case-10 | pass→pass | 3,793 | 2,667 | -30% | 1 | 1 | 0% | 490 | 1,153 | +135% | 0 | 0 | — |
case-11 | fail→pass | 9,999 | 7,161 | -28% | 1 | 1 | 0% | 1,504 | 1,785 | +19% | 0 | 0 | — |
case-12 | fail→pass | 21,230 | 3,460 | -84% | 1 | 1 | 0% | 3,600 | 1,241 | -66% | 0 | 0 | — |
case-13 | pass→pass | 5,775 | 2,828 | -51% | 1 | 1 | 0% | 908 | 1,156 | +27% | 0 | 0 | — |
case-14 | fail→pass | 9,730 | 2,023 | -79% | 1 | 1 | 0% | 1,421 | 985 | -31% | 0 | 0 | — |
case-15 | fail→fail | 6,014 | 2,860 | -52% | 1 | 1 | 0% | 843 | 1,072 | +27% | 0 | 0 | — |
case-16 | pass→pass | 14,460 | 2,219 | -85% | 1 | 1 | 0% | 2,021 | 991 | -51% | 0 | 0 | — |
case-17 | fail→pass | 16,974 | 3,180 | -81% | 1 | 1 | 0% | 2,635 | 1,189 | -55% | 0 | 0 | — |
case-18 | fail→pass | 14,165 | 1,701 | -88% | 1 | 1 | 0% | 2,145 | 924 | -57% | 0 | 0 | — |
case-19 | fail→pass | 6,906 | 2,626 | -62% | 1 | 1 | 0% | 1,107 | 1,136 | +3% | 0 | 0 | — |
case-20 | pass→pass | 7,655 | 5,842 | -24% | 1 | 1 | 0% | 1,110 | 1,574 | +42% | 0 | 0 | — |
case-21 | fail→pass | 11,702 | 7,047 | -40% | 1 | 1 | 0% | 1,797 | 1,716 | -5% | 0 | 0 | — |
case-22 | pass→fail | 12,376 | 5,844 | -53% | 1 | 1 | 0% | 1,697 | 1,591 | -6% | 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, and 19 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +41 percentage points is the difference between those two pass rates over the 19 comparable cases. 1 case got worse with the skill loaded, and it is 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.