Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when an external event thesis needs market confirmation before commitment.
.claude/skills/superior-trade-catalyst-confirmation/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | -23% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -54% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -14% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -7% | 0% |
Use this when the user ties a strategy idea to an external event (CPI, Fed, court rulings, ETF, earnings, protocol upgrade, sports injuries) and wants Polymarket behavior to confirm it.
POST /v3/markets/search.Moderate fit. Filled TradeTick backtests can validate whether price reacted and how quickly after timestamps. They do not validate the news source itself or guarantee it remains relevant in future sessions.
This is an archetype template: it should be converted into custom NautilusTrader code and tuned by the user and engine constraints.
Enter only when the expected directional catalyst outcome appears and a filled-trade confirmation move occurs after the catalyst window.
confirmation_move in the right direction within confirmation_window_ticks.exit_after_ticks, on reversal, or near market end.json{ "catalyst_time": "2026-06-12T18:00:00Z", "confirmation_move": 0.025, "confirmation_window_ticks": 20, "order_size": 10, "exit_after_ticks": 60 }
| Knob | Effect | |---|---| | confirmation_move | Higher = wait for stronger post-catalyst movement. | | confirmation_window_ticks | Lower = tighter catalyst-response window. | | exit_after_ticks | Higher = hold longer for delayed event drift. |
"This is a catalyst-led archetype, not a guaranteed edge. It runs on filled TradeTick history and exact market slugs, then becomes a starting point for custom Nautilus strategy generation."
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 32,282 | 19,059 | -41% | 1 | 1 | 0% | 6,254 | 3,958 | -37% | 0 | 0 | — |
case-02 | fail→fail | 44,367 | 9,108 | -79% | 1 | 1 | 0% | 3,923 | 2,139 | -45% | 0 | 0 | — |
case-03 | fail→fail | 22,147 | 14,717 | -34% | 1 | 1 | 0% | 3,179 | 3,353 | +5% | 0 | 0 | — |
case-04 | pass→pass | 28,284 | 33,112 | +17% | 1 | 1 | 0% | 4,783 | 5,954 | +24% | 0 | 0 | — |
case-05 | pass→pass | 22,281 | 23,187 | +4% | 1 | 1 | 0% | 4,199 | 5,102 | +22% | 0 | 0 | — |
case-06 | pass→pass | 31,252 | 23,179 | -26% | 1 | 1 | 0% | 4,690 | 4,851 | +3% | 0 | 0 | — |
case-07 | pass→pass | 18,570 | 20,834 | +12% | 1 | 1 | 0% | 3,021 | 4,055 | +34% | 0 | 0 | — |
case-08 | fail→pass | 21,426 | 11,334 | -47% | 1 | 1 | 0% | 3,162 | 2,422 | -23% | 0 | 0 | — |
case-09 | fail→pass | 16,556 | 12,059 | -27% | 1 | 1 | 0% | 2,886 | 2,602 | -10% | 0 | 0 | — |
case-10 | fail→fail | 20,867 | 12,063 | -42% | 1 | 1 | 0% | 2,984 | 2,613 | -12% | 0 | 0 | — |
case-16 | fail→pass | 16,245 | 2,950 | -82% | 1 | 1 | 0% | 2,226 | 1,035 | -54% | 0 | 0 | — |
case-11 | fail→pass | 16,735 | 11,170 | -33% | 1 | 1 | 0% | 2,442 | 2,106 | -14% | 0 | 0 | — |
case-12 | fail→pass | 16,489 | 10,237 | -38% | 1 | 1 | 0% | 2,446 | 2,264 | -7% | 0 | 0 | — |
case-13 | fail→pass | 18,703 | 14,975 | -20% | 1 | 1 | 0% | 2,679 | 2,601 | -3% | 0 | 0 | — |
case-14 | fail→fail | 13,053 | 12,615 | -3% | 1 | 1 | 0% | 2,065 | 2,368 | +15% | 0 | 0 | — |
case-15 | pass→pass | 11,037 | 5,795 | -47% | 1 | 1 | 0% | 1,687 | 1,507 | -11% | 0 | 0 | — |
case-17 | pass→pass | 12,685 | 6,903 | -46% | 1 | 1 | 0% | 2,024 | 1,810 | -11% | 0 | 0 | — |
case-18 | pass→pass | 14,822 | 15,814 | +7% | 1 | 1 | 0% | 2,280 | 2,859 | +25% | 0 | 0 | — |
case-19 | pass→pass | 14,681 | 15,528 | +6% | 1 | 1 | 0% | 2,198 | 2,777 | +26% | 0 | 0 | — |
case-20 | pass→pass | 11,764 | 8,912 | -24% | 1 | 1 | 0% | 1,690 | 1,701 | +1% | 0 | 0 | — |
case-21 | pass→pass | 8,779 | 4,868 | -45% | 1 | 1 | 0% | 1,483 | 1,451 | -2% | 0 | 0 | — |
case-22 | fail→fail | 14,350 | 10,625 | -26% | 1 | 1 | 0% | 2,700 | 2,512 | -7% | 0 | 0 | — |
case-23 | fail→pass | 16,072 | 16,929 | +5% | 1 | 1 | 0% | 2,324 | 3,135 | +35% | 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. 23 cases were attempted. The headline lift of +30 percentage points is the difference between those two pass rates over the 23 comparable cases.
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.