SuperAPI AI Directory

htx/ta-master

General & other

View repository

Updated 10-16 07:20

About this skill

Technical Analysis Master

Layer 2 orchestration skill, no API key required. Composes 6 L1 skills into a three-pillar score and provides a 0-100 composite read.

Compliance disclaimer: the score is a mechanical algorithmic output and does not constitute investment advice. Markets carry risk; all decisions are made by the user.

Three-pillar scoring framework

Pillar Weight Data source Output
Price/Volume 50% htx/technical-analysis (51 indicators + 12 patterns + divergences) 0-100 sub-score
Derivatives 30% funding-rate / oi-tracker / liquidation-stream / elite-positioning / mark-price 0-100 sub-score
BTC cycle 20% htx/technical-analysis cycle.py (BTC-USDT only) 0-100 sub-score

For non-BTC assets the weights are auto-redistributed to price/volume 62.5% + derivatives 37.5%.

Composite score interpretation

Composite score Label
≥ 70 STRONG BULLISH
55-70 MILD BULLISH
45-55 NEUTRAL
30-45 MILD BEARISH
< 30 STRONG BEARISH

Workflow

Full version (BTC, three pillars)

# 1. Pull daily klines (cycle pillar needs 350+ daily klines)
htx-cli spot-market kline -p symbol=btcusdt -p period=1day -p size=400 | jq '.data' > /tmp/btc1d.json
htx-cli spot-market kline -p symbol=btcusdt -p period=4hour -p size=300 | jq '.data' > /tmp/btc4h.json

# 2. Compute price/volume features → pv.json
python -c "
import json
import indicators, patterns
df = indicators._df(json.load(open('/tmp/btc4h.json')))
out = {
    'rsi': indicators.rsi(df)['rsi'].iloc[-1],
    'macd_hist': indicators.macd(df)['macd'].iloc[-1],
    'ema_fast': indicators.ema(df, periods=(20,))['ema20'].iloc[-1],
    'ema_slow': indicators.ema(df, periods=(60,))['ema60'].iloc[-1],
    'adx': indicators.adx(df)['adx'].iloc[-1],
    'divergence': str(indicators.divergence(df)['divergence'].iloc[-1]),
    'patterns_bullish_count': sum(1 for p in patterns.scan(df) if p.startswith('bull') or p in ('three-soldiers','inverted-hammer')),
    'patterns_bearish_count': sum(1 for p in patterns.scan(df) if p.startswith('bear') or p in ('three-crows','shooting-star','hanging-man')),
}
json.dump(out, open('/tmp/pv.json', 'w'))
"

# 3. Pull derivatives → deriv.json
htx-cli funding-rate current -p contract_code=BTC-USDT --json > /tmp/fr.json
htx-cli oi-tracker history -p contract_code=BTC-USDT -p period=60min -p size=24 --json > /tmp/oi.json
htx-cli liquidation-stream recent -p contract=BTC-USDT --json > /tmp/liq.json
htx-cli elite-positioning ratio -p contract_code=BTC-USDT --json > /tmp/elite.json
htx-cli mark-price basis -p contract_code=BTC-USDT --json > /tmp/basis.json
# Then aggregate into deriv.json (see references/derivatives-features.md)

# 4. Compute cycle pillar → cycle.json (BTC only)
python scripts/cycle.py all --kline /tmp/btc1d.json > /tmp/cycle_raw.json
# Extract key fields into cycle.json: ahr999 / mayer / pi_cycle_signal / rainbow_band

# 5. Three-pillar combined score
python scripts/scoring.py --pricevol /tmp/pv.json --derivatives /tmp/deriv.json --cycle /tmp/cycle.json

Simple version (non-BTC, two pillars)

python scripts/scoring.py --pricevol pv.json --derivatives deriv.json

Example output (BTC full three-pillar):

{
  "composite": {
    "composite": 62.4,
    "label": "MILD BULLISH",
    "weights": {"pv": 0.5, "deriv": 0.3, "cycle": 0.2}
  },
  "pillars": {
    "price_volume": {
      "score": 65.0,
      "notes": ["RSI 58.3 bullish", "MACD hist > 0", "EMA fast > slow (uptrend)"]
    },
    "derivatives": {
      "score": 55.0,
      "notes": ["Funding 0.012% — bullish bias", "Elite L/S 1.18 — leaning long"]
    },
    "cycle": {
      "score": 70.0,
      "notes": ["AHR999 1.05 — DCA zone", "Mayer 1.32 — fair", "Rainbow: HODL!"]
    }
  }
}

Scoring details

See scripts/scoring.py for the source. Each pillar lists its contributing items and notes.

Price/volume pillar items

Signal Add / subtract
RSI > 70 (overbought) -10
RSI > 55 (bullish-leaning) +5
RSI < 30 (oversold) +10
RSI < 45 (bearish-leaning) -5
MACD hist > 0 / < 0 ±7
EMA fast above / below slow ±8
ADX > 25 strong-trend amplifier (no direct score)
Bullish regular divergence (bull_reg) +12
Bearish regular divergence (bear_reg) -12
Bullish / bearish hidden divergence ±6
Each bullish pattern +4
Each bearish pattern -4

Derivatives pillar items

Signal Add / subtract
Funding > 0.05% -12 (longs overheated)
Funding < -0.05% +12 (shorts overheated)
OI 24h +15% -8 (squeeze risk)
OI 24h -10% -6 (capital exiting)
Elite L/S > 1.5 +10 (smart money long)
Elite L/S < 0.7 -10 (smart money short)
1h long-liquidation share > 80% +8 (bottom signal)
1h short-liquidation share > 80% -8 (top signal)
Basis deviation ±0.5% ±4

BTC cycle pillar items

Signal Add / subtract
AHR999 < 0.45 +20 (bottom-fishing zone)
AHR999 > 1.6 -20 (bubble warning)
Pi Cycle TOP triggered -25 WARNING
Mayer < 1 +8
Mayer > 2.4 -12
Rainbow in Fire Sale / BUY / Accumulate +10
Rainbow in FOMO / Sell / Bubble -10

Data gaps (honest disclosure)

ta-master does not cover the following indicators (HTX provides no native endpoint; paid data sources are required):

Gap Source
MVRV / NUPL / SOPR Glassnode (on-chain)
Hash Ribbon / miner hashrate Mempool.space
LTH/STH supply Glassnode
All-account long/short ratio (retail breakdown) HTX only provides the elite breakdown
Taker active buy/sell volume Not provided by HTX
Liquidation heatmap density We aggregate from the local liquidation order stream; density is lower than Coinglass

Future option: integrate via --external-source glassnode etc.

Dependent skills

Before installing this skill, ensure the following are installed:

  • htx/spot-market (kline source)
  • htx/futures-market (kline source)
  • htx/technical-analysis (indicator computation engine)
  • htx/funding-rate
  • htx/oi-tracker
  • htx/liquidation-stream
  • htx/elite-positioning
  • htx/mark-price

Installation

curl -fsSL https://github.com/htx-exchange/htx-skills-hub/releases/latest/download/install.sh | bash -s -- ta-master

Typical questions

  • "How does BTC look right now combining technicals + derivatives + cycle?"
  • "ETH 4H composite score"
  • "Market-wide scan, give me coins with ta-master score > 70"
  • "Judge BTC right now using AHR999 + funding rate + RSI together"