Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Detect and classify candlestick patterns from ingested OHLCV data
.claude/skills/ruvnet-market-pattern/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-14 | ✗→✓ | ▲ Improved | -42% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 59% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -25% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -34% | 0% |
Scan ingested OHLCV data for known candlestick patterns, classify them by type and reliability, and store for future reference.
When you need to identify candlestick patterns (doji, hammer, engulfing, head-shoulders, etc.) in market data. Requires data to be ingested first via market-ingest.
mcp__plugin_ruflo-core_ruflo__memory_search (or memory_list) on the market-data namespace to retrieve normalized OHLCV data for the symbol and period. The memory_* tool family routes by namespace; the agentdb_hierarchical-* family does NOT (it routes by tier), so use memory_* for namespaced reads.mcp__plugin_ruflo-core_ruflo__agentdb_pattern-store with type: 'market-pattern'. Don't pass a namespace arg — ReasoningBank routes it; on bridge unavailability the fallback writes to the reserved pattern namespace with controller: 'memory-store-fallback' (see ruflo-agentdb ADR-0001).mcp__plugin_ruflo-core_ruflo__memory_store --namespace market-patterns — this DOES respect the market-patterns namespace because memory_* is namespace-routed.bashnpx @claude-flow/cli@latest memory search --query "bullish reversal patterns" --namespace market-patterns npx @claude-flow/cli@latest memory store --key "pattern-AAPL-2026-05-04-doji" --value '{...}' --namespace market-patterns
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-13 | pass→pass | 6,908 | 2,335 | -66% | 1 | 1 | 0% | 1,428 | 1,073 | -25% | 0 | 0 | — |
case-14 | fail→pass | 7,518 | 1,554 | -79% | 1 | 1 | 0% | 1,520 | 876 | -42% | 0 | 0 | — |
case-01 | fail→fail | 16,788 | 5,122 | -69% | 1 | 1 | 0% | 3,086 | 886 | -71% | 0 | 0 | — |
case-02 | fail→fail | 14,844 | 9,313 | -37% | 1 | 1 | 0% | 2,851 | 981 | -66% | 0 | 0 | — |
case-03 | fail→fail | 15,387 | 4,788 | -69% | 1 | 1 | 0% | 3,297 | 972 | -71% | 0 | 0 | — |
case-04 | pass→pass | 17,210 | 15,658 | -9% | 1 | 1 | 0% | 3,657 | 3,123 | -15% | 0 | 0 | — |
case-05 | pass→fail | 14,578 | 4,597 | -68% | 1 | 1 | 0% | 2,104 | 954 | -55% | 0 | 0 | — |
case-06 | pass→pass | 1,445 | 4,969 | +244% | 1 | 1 | 0% | 253 | 1,535 | +507% | 0 | 0 | — |
case-07 | fail→pass | 9,808 | 3,914 | -60% | 1 | 1 | 0% | 1,741 | 1,418 | -19% | 0 | 0 | — |
case-08 | pass→fail | 7,187 | 11,584 | +61% | 1 | 1 | 0% | 1,470 | 1,675 | +14% | 0 | 0 | — |
case-09 | pass→pass | 3,973 | 4,962 | +25% | 1 | 1 | 0% | 731 | 1,659 | +127% | 0 | 0 | — |
case-10 | pass→pass | 6,064 | 4,404 | -27% | 1 | 1 | 0% | 1,126 | 1,395 | +24% | 0 | 0 | — |
case-11 | pass→pass | 5,793 | 5,239 | -10% | 1 | 1 | 0% | 1,065 | 1,597 | +50% | 0 | 0 | — |
case-12 | pass→pass | 13,074 | 5,116 | -61% | 1 | 1 | 0% | 2,375 | 1,527 | -36% | 0 | 0 | — |
case-15 | fail→pass | 5,236 | 3,880 | -26% | 1 | 1 | 0% | 872 | 1,386 | +59% | 0 | 0 | — |
case-16 | fail→pass | 7,331 | 2,078 | -72% | 1 | 1 | 0% | 1,357 | 1,015 | -25% | 0 | 0 | — |
case-17 | fail→pass | 13,403 | 5,242 | -61% | 1 | 1 | 0% | 2,499 | 1,644 | -34% | 0 | 0 | — |
case-18 | fail→pass | 4,257 | 1,574 | -63% | 1 | 1 | 0% | 820 | 885 | +8% | 0 | 0 | — |
case-19 | fail→pass | 5,296 | 2,661 | -50% | 1 | 1 | 0% | 1,209 | 1,138 | -6% | 0 | 0 | — |
case-20 | pass→pass | 9,168 | 6,795 | -26% | 1 | 1 | 0% | 1,561 | 1,792 | +15% | 0 | 0 | — |
case-21 | pass→pass | 3,219 | 3,425 | +6% | 1 | 1 | 0% | 588 | 1,299 | +121% | 0 | 0 | — |
case-22 | pass→pass | 3,330 | 9,620 | +189% | 1 | 1 | 0% | 551 | 1,531 | +178% | 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 17 counted toward the lift figure. The other 5 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 +23 percentage points is the difference between those two pass rates over the 17 comparable cases. 4 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.