Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when a Polymarket market probability changes as the resolution deadline approaches.
.claude/skills/superior-trade-deadline-drift/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -39% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -34% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -35% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -28% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 5% | 0% |
Use this for before-date contracts, monthly/weekly threshold markets, election timing markets, or requests where time remaining is the core thesis.
POST /v3/markets/search.market_end from market metadata.Moderate fit. Filled TradeTick data can show whether price drifted in line with deadline pressure and whether exits would usually occur before binary resolution.
Limit: this does not produce true fair-probability forecasts; it only tests historical price behavior around deadlines.
Enter when drift aligns with time decay or deadline acceleration. Exit before settlement unless the user explicitly asks for binary exposure into resolution.
on_trade_tick.market_end.min_days_to_end and max_days_to_end.exit_buffer_hours cutoff.json{ "market_end": "2026-07-01T00:00:00Z", "min_days_to_end": 2, "max_days_to_end": 21, "drift_threshold": 0.015, "order_size": 10, "exit_buffer_hours": 12 }
| Knob | Effect | |---|---| | min_days_to_end | Avoids entering too close to settlement. | | max_days_to_end | Avoids entering when the deadline is still distant. | | drift_threshold | Higher values require stronger deadline repricing. | | exit_buffer_hours | Larger buffers reduce resolution timing risk. |
"This strategy trades deadline pressure, not certainty. I will backtest on filled TradeTicks to check whether the historical price drift is consistent and tradable, then suggest only cautious position sizing."
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 16,703 | 5,908 | -65% | 1 | 1 | 0% | 2,821 | 1,730 | -39% | 0 | 0 | — |
case-02 | fail→fail | 26,105 | 6,559 | -75% | 1 | 1 | 0% | 1,280 | 1,812 | +42% | 0 | 0 | — |
case-03 | fail→pass | 16,006 | 7,299 | -54% | 1 | 1 | 0% | 2,980 | 1,956 | -34% | 0 | 0 | — |
case-04 | pass→fail | 24,730 | 23,400 | -5% | 1 | 1 | 0% | 4,199 | 3,592 | -14% | 0 | 0 | — |
case-05 | pass→pass | 20,336 | 17,019 | -16% | 1 | 1 | 0% | 3,477 | 3,566 | +3% | 0 | 0 | — |
case-06 | pass→fail | 21,790 | 30,481 | +40% | 1 | 1 | 0% | 4,103 | 6,730 | +64% | 0 | 0 | — |
case-07 | fail→pass | 7,750 | 2,039 | -74% | 1 | 1 | 0% | 1,336 | 872 | -35% | 0 | 0 | — |
case-08 | fail→pass | 17,404 | 10,350 | -41% | 1 | 1 | 0% | 2,924 | 2,106 | -28% | 0 | 0 | — |
case-09 | fail→pass | 16,241 | 13,998 | -14% | 1 | 1 | 0% | 2,695 | 2,839 | +5% | 0 | 0 | — |
case-10 | fail→pass | 13,207 | 5,274 | -60% | 1 | 1 | 0% | 1,949 | 1,490 | -24% | 0 | 0 | — |
case-11 | pass→pass | 13,288 | 6,062 | -54% | 1 | 1 | 0% | 1,983 | 1,526 | -23% | 0 | 0 | — |
case-20 | fail→pass | 14,502 | 3,710 | -74% | 1 | 1 | 0% | 2,173 | 1,213 | -44% | 0 | 0 | — |
case-12 | pass→pass | 17,130 | 12,475 | -27% | 1 | 1 | 0% | 2,620 | 2,349 | -10% | 0 | 0 | — |
case-13 | fail→pass | 9,979 | 6,837 | -31% | 1 | 1 | 0% | 1,478 | 1,562 | +6% | 0 | 0 | — |
case-14 | pass→pass | 11,216 | 3,130 | -72% | 1 | 1 | 0% | 1,596 | 1,116 | -30% | 0 | 0 | — |
case-15 | pass→pass | 9,719 | 5,823 | -40% | 1 | 1 | 0% | 1,529 | 1,420 | -7% | 0 | 0 | — |
case-16 | fail→pass | 11,419 | 6,858 | -40% | 1 | 1 | 0% | 1,778 | 1,562 | -12% | 0 | 0 | — |
case-17 | pass→pass | 19,656 | 11,275 | -43% | 1 | 1 | 0% | 2,942 | 2,347 | -20% | 0 | 0 | — |
case-18 | pass→pass | 13,738 | 13,147 | -4% | 1 | 1 | 0% | 1,863 | 2,635 | +41% | 0 | 0 | — |
case-19 | fail→pass | 12,717 | 4,398 | -65% | 1 | 1 | 0% | 1,754 | 1,242 | -29% | 0 | 0 | — |
case-21 | fail→pass | 12,350 | 2,083 | -83% | 1 | 1 | 0% | 1,793 | 964 | -46% | 0 | 0 | — |
case-22 | fail→pass | 21,845 | 9,646 | -56% | 1 | 1 | 0% | 3,150 | 2,004 | -36% | 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 21 counted toward the lift figure. The other 1 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 +45 percentage points is the difference between those two pass rates over the 21 comparable cases. 2 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.