Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Smart Money Concepts (SMC) and ICT-style chart analysis skill. Analyzes any market with one command - fetches real data, runs multi-timeframe analysis, identifies trade setups with entry/stop/targets. Use when user asks to "analyze [SYMBOL]", wants SMC/ICT analysis, asks about liquidity sweeps, order blocks, fair value gaps, or market structure, or wants trade setup recommendations.
.claude/skills/aiskillstore-smc-chart-analysis/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 78% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 141% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 32% | 0% |
| case-05 | ✓→✓ | = Same ✓ | 120% | 0% |
AI-powered Smart Money Concepts analysis with actionable trade setups.
┌─────────────────────────────────────────────────────────────┐
│ 1. FETCH DATA │
│ Call web app API: localhost:3001/api/smc-analyze │
│ → Returns: structure, liquidity, FVGs, sweeps │
└─────────────────────────┬───────────────────────────────────┘
│
▼
┌─────────────────────────────────────────────────────────────┐
│ 2. AI INTERPRETATION (Claude) │
│ Analyze the mechanical data using ICT methodology │
│ → Determine bias, identify setups, assess confluence │
└─────────────────────────┬───────────────────────────────────┘
│
▼
┌─────────────────────────────────────────────────────────────┐
│ 3. TRADE SETUP OUTPUT │
│ • Direction (long/short) │
│ • Entry zone with reasoning │
│ • Stop loss with reasoning │
│ • Targets with R:R ratios │
│ • Setup grade (A+/A/B/C) │
│ • Narrative explaining the trade thesis │
└─────────────────────────────────────────────────────────────┘When a user asks to analyze a market (e.g., "analyze BTC", "what's the setup on ETH"):
bashcurl "http://localhost:3001/api/smc-analyze?symbol=BTC/USDT&timeframe=1h&style=intraday"
Or for multiple timeframes:
bash# HTF (4H) for bias curl "http://localhost:3001/api/smc-analyze?symbol=BTC/USDT&timeframe=4h&style=swing" # LTF (15m) for entry curl "http://localhost:3001/api/smc-analyze?symbol=BTC/USDT&timeframe=15m&style=scalp"
Take the mechanical data and provide AI-powered analysis following the ICT methodology:
Format the response as:
## [SYMBOL] SMC Analysis
### Bias: [BULLISH/BEARISH]
[Explain why based on structure, premium/discount zone, and liquidity]
### Current Setup: [GRADE]
**Direction:** [LONG/SHORT]
**Entry Zone:** $XX,XXX - $XX,XXX
- Reasoning: [Why this zone - FVG? OB? Sweep level?]
**Stop Loss:** $XX,XXX
- Reasoning: [Below sweep low / Above sweep high]
- Risk: X.X%
**Targets:**
| Target | Price | R:R | Reasoning |
|--------|-------|-----|-----------|
| TP1 | $XX,XXX | 1:2 | [Nearest liquidity] |
| TP2 | $XX,XXX | 1:3 | [Major BSL/SSL] |
| TP3 | $XX,XXX | 1:5 | [HTF liquidity] |
### Confluence Factors
- [ ] HTF bias aligned
- [ ] Price in discount/premium
- [ ] Recent sweep occurred
- [ ] ChoCH confirmed
- [ ] FVG/OB present at entry
- [ ] In killzone (if applicable)
### Warnings
- [Any concerns about the setup]
### Narrative
[2-3 sentence explanation of the trade thesis in plain English]
---
*This is analysis, not financial advice. Always manage risk.*The web app at localhost:3001 provides:
To start the web app:
bashcd /Users/abuusama/projects/bots/signals && bun dev
The web app shows the mechanical data. This skill provides the AI interpretation.
Parameters: | Param | Default | Options | |-------|---------|---------| | symbol | BTC/USDT | BTC/USDT, ETH/USDT, SOL/USDT | | timeframe | 1h | 1m, 5m, 15m, 1h, 4h, 1d, 1w | | style | intraday | scalp, intraday, swing, position |
Response:
json{ "symbol": "BTC/USDT", "timeframe": "1h", "style": "intraday", "currentPrice": 88338.29, "structure": { "bias": "BULLISH", "priceZone": "PREMIUM", "range": { "high": 89399.97, "low": 84450.01 } }, "liquidity": { "bsl": [{ "price": 92754.00, "strength": "strong" }], "ssl": [{ "price": 84450.01, "strength": "strong" }] }, "inefficiencies": { "fvgs": [{ "type": "bull", "high": 88049.89, "low": 87946.96 }] }, "orderFlow": { "recentSwings": [...], "structureBreaks": [{ "type": "CHOCH", "direction": "up", "price": 88339.27 }] }, "atr": 132.75 }
| Session | Time | Best For | |---------|------|----------| | London Open | 02:00-05:00 | Manipulation moves | | NY AM | 07:00-10:00 | Highest probability | | London Close | 10:00-12:00 | Reversal setups |
User: "analyze BTC"
bashcurl "http://localhost:3001/api/smc-analyze?symbol=BTC/USDT&timeframe=1h"
## BTC/USDT SMC Analysis
### Bias: BULLISH
Price is currently at $88,338 in the PREMIUM zone of the 1H range ($84,450 - $89,399).
However, we just had a CHOCH to the upside at $88,339, suggesting bullish continuation
despite being in premium.
### Current Setup: B+
**Direction:** LONG (on pullback)
**Entry Zone:** $87,946 - $88,049
- Reasoning: Unfilled bullish FVG from recent impulse move
**Stop Loss:** $87,500
- Reasoning: Below the recent BOS level at $87,514
- Risk: 0.95%
**Targets:**
| Target | Price | R:R | Reasoning |
|--------|-------|-----|-----------|
| TP1 | $89,400 | 1:1.7 | Range high / BSL |
| TP2 | $90,634 | 1:3.2 | Next BSL level |
| TP3 | $92,754 | 1:5.7 | Major BSL |
### Confluence Factors
- [x] ChoCH confirmed bullish
- [x] FVG present at entry zone
- [ ] Not in killzone currently
- [ ] Price in premium (slight concern)
### Warnings
- Price is in premium zone - ideal longs are from discount
- Consider waiting for deeper pullback to $87,500 area
### Narrative
BTC showed a change of character to the upside, breaking recent structure.
The setup suggests buying a pullback into the unfilled FVG around $88,000,
with stops below recent structure and targets at the BSL levels 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 | 15,192 | 6,116 | -60% | 1 | 1 | 0% | 2,479 | 2,577 | +4% | 0 | 0 | — |
case-07 | fail→pass | 9,877 | 4,191 | -58% | 1 | 1 | 0% | 1,592 | 2,841 | +78% | 0 | 0 | — |
case-02 | fail→fail | 16,702 | 15,883 | -5% | 1 | 1 | 0% | 2,572 | 4,994 | +94% | 0 | 0 | — |
case-03 | fail→fail | 16,978 | 5,637 | -67% | 1 | 1 | 0% | 2,617 | 2,587 | -1% | 0 | 0 | — |
case-04 | fail→pass | 17,563 | 8,634 | -51% | 1 | 1 | 0% | 3,048 | 3,870 | +27% | 0 | 0 | — |
case-05 | pass→pass | 8,715 | 5,797 | -33% | 1 | 1 | 0% | 1,405 | 3,085 | +120% | 0 | 0 | — |
case-06 | pass→pass | 7,717 | 1,908 | -75% | 1 | 1 | 0% | 1,245 | 2,581 | +107% | 0 | 0 | — |
case-08 | pass→pass | 7,449 | 2,145 | -71% | 1 | 1 | 0% | 1,229 | 2,563 | +109% | 0 | 0 | — |
case-09 | pass→pass | 14,170 | 6,706 | -53% | 1 | 1 | 0% | 2,173 | 3,376 | +55% | 0 | 0 | — |
case-10 | pass→pass | 17,160 | 9,519 | -45% | 1 | 1 | 0% | 2,595 | 3,819 | +47% | 0 | 0 | — |
case-11 | pass→pass | 8,205 | 5,551 | -32% | 1 | 1 | 0% | 1,231 | 3,029 | +146% | 0 | 0 | — |
case-12 | pass→pass | 16,577 | 9,328 | -44% | 1 | 1 | 0% | 2,395 | 3,520 | +47% | 0 | 0 | — |
case-13 | pass→pass | 9,427 | 5,370 | -43% | 1 | 1 | 0% | 1,573 | 3,155 | +101% | 0 | 0 | — |
case-14 | fail→pass | 7,499 | 1,820 | -76% | 1 | 1 | 0% | 1,079 | 2,596 | +141% | 0 | 0 | — |
case-15 | fail→pass | 11,205 | 2,177 | -81% | 1 | 1 | 0% | 1,961 | 2,584 | +32% | 0 | 0 | — |
case-16 | pass→pass | 10,353 | 2,717 | -74% | 1 | 1 | 0% | 1,648 | 2,731 | +66% | 0 | 0 | — |
case-22 | pass→pass | 9,593 | 7,593 | -21% | 1 | 1 | 0% | 1,806 | 3,631 | +101% | 0 | 0 | — |
case-17 | pass→pass | 10,249 | 6,452 | -37% | 1 | 1 | 0% | 1,663 | 3,231 | +94% | 0 | 0 | — |
case-18 | pass→pass | 8,911 | 6,691 | -25% | 1 | 1 | 0% | 1,450 | 3,336 | +130% | 0 | 0 | — |
case-19 | pass→pass | 13,541 | 2,978 | -78% | 1 | 1 | 0% | 2,116 | 2,717 | +28% | 0 | 0 | — |
case-20 | pass→pass | 13,488 | 12,987 | -4% | 1 | 1 | 0% | 2,633 | 4,873 | +85% | 0 | 0 | — |
case-21 | pass→pass | 8,617 | 8,975 | +4% | 1 | 1 | 0% | 1,620 | 3,923 | +142% | 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 20 counted toward the lift figure. The other 2 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 +18 percentage points is the difference between those two pass rates over the 20 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.