SKILL.md
Technical Analysis
Read the chart systematically — don't cherry-pick indicators that confirm a bias.
Prerequisites
No hard dependencies. Works best with Groww MCP or yfinance for live data, but can analyze any stock if you provide prices manually.
Data Collection
When data tools are available, fetch in this order:
- Resolve symbol:
searchstockandotherssymbolon Groww - Live quote:
getquotesand_depthfor current price, volume, day range - Indicators:
gethistoricaltechnical_indicatorsfor SMA, EMA, RSI, MACD, Bollinger, ADX - Candles:
fetchhistoricalcandle_datawithinterval=1dfor at least 200 days - Long history (if needed):
yf.download("SYMBOL.NS", period="2y")via yfinance
Step 1: Trend Identification
- Primary trend (weekly): Higher highs & higher lows = uptrend. Lower highs & lower lows = downtrend.
- Secondary trend (daily): Current swing direction within the primary trend.
- Moving average alignment:
- Price > SMA20 > SMA50 > SMA200 = strong uptrend - Price < SMA20 < SMA50 < SMA200 = strong downtrend - Mixed = transitional / choppy
- Cross signals: Golden cross (SMA50 > SMA200) = bullish. Death cross = bearish.
Step 2: Support & Resistance
- Horizontal S/R: Recent swing highs and lows with at least 2 touches
- Round numbers: Psychological levels (Rs.1,000, Rs.1,500, Rs.2,000)
- Volume clusters: Zones where heavy trading occurred (act as magnets)
- Dynamic S/R: SMA20 (short-term), SMA50 (medium), SMA200 (major)
- Gap zones: Unfilled gaps act as support (gap up) or resistance (gap down)
Step 3: Volume Analysis
| Price Action | Volume | Interpretation |
|---|---|---|
| Rising price | Rising volume | Strong move — trend healthy |
| Rising price | Falling volume | Weak rally — suspect, may reverse |
| Falling price | Rising volume | Strong selling — trend acceleration |
| Falling price | Falling volume | Selling exhaustion — bounce possible |
| Spike volume at support | — | Potential capitulation / reversal |
| Spike volume at resistance | — | Breakout attempt or distribution |
Compare current volume to 20-day average. >1.5x = significant.
Step 4: Indicator Dashboard
Pull via gethistoricaltechnical_indicators:
| Indicator | Bullish | Bearish | Neutral |
|---|---|---|---|
| RSI(14) | >50, rising from <30 | <50, falling from >70 | 40-60, flat |
| MACD | Line > Signal, histogram growing | Line < Signal, histogram shrinking | Near zero line |
| Bollinger Bands | Price bouncing off lower band | Price rejected at upper band | Walking the band |
| ADX | >25 with +DI > -DI | >25 with -DI > +DI | <20 (no trend) |
| Stochastic | %K crosses above %D below 20 | %K crosses below %D above 80 | Mid-range |
Count bullish vs bearish signals to gauge overall bias.
Step 5: Probabilistic Scenarios
Always present 2-3 scenarios with rough probabilities. Assign higher probability to the scenario with more technical evidence.
SCENARIO A (Bullish) — XX% probability
- Trigger: [specific price/indicator condition]
- Target: Rs.XXX (nearest resistance / Fib level)
- Invalidation: Close below Rs.XXX
SCENARIO B (Bearish) — XX% probability
- Trigger: [specific condition]
- Target: Rs.XXX (nearest support)
- Invalidation: Close above Rs.XXX
SCENARIO C (Sideways) — XX% probability
- Range: Rs.XXX to Rs.XXX
- Duration estimate: X days/weeks
- Breakout signal: [what to watch]Probabilities should sum to ~100%. Base them on: trend alignment, indicator consensus, volume confirmation, and S/R proximity.