Htx/technical Analysis

Technical Analysis — Indicator Engine v3

Local computation engine, no API key required. Pulls data from HTX kline endpoints and computes indicators / patterns / cycles locally with numpy/pandas.

Compliance disclaimer: this skill provides raw indicator values and does not embed any strategy recommendations or trading advice. All decisions are made by the user based on their own risk tolerance.

Capability overview

Quick start

Pull klines + compute indicators

# 1. Pull BTC/USDT 4-hour klines
htx-cli spot-market kline -p symbol=btcusdt -p period=4hour -p size=300 \
  | jq '.data' > /tmp/btc4h.json

# 2. Compute RSI
python scripts/indicators.py rsi --kline /tmp/btc4h.json --params 14
# → {"rsi": 62.4, "ts": 1779000000000}

# 3. Compute MACD (default 12,26,9)
python scripts/indicators.py macd --kline /tmp/btc4h.json
# → {"dif": 320.1, "dea": 245.3, "macd": 149.6, "ts": ...}

# 4. Scan all candlestick patterns
python scripts/patterns.py scan --kline /tmp/btc4h.json
# → {"patterns": ["doji", "bull-engulf"], "ts": ...}

# 5. BTC cycle one-shot full run
python scripts/cycle.py all --kline /tmp/btc1d.json

List all indicators

python scripts/indicators.py list
# → ["ma", "ema", "rsi", "macd", "supertrend", ...]

Command reference

See references/:

• references/indicators.md — parameters / return fields / formulas for the 51 technical indicators
• references/patterns.md — judgment rules and typical scenarios for the 12 candlestick patterns
• references/cycle.md — formulas and interpretation ranges for the 5 BTC cycle indicators
• references/divergence.md — automatic divergence detection algorithm and usage notes

Automatic divergence detection

python scripts/indicators.py divergence --kline /tmp/btc4h.json --params 14
# → {"divergence": "bull_reg", ...}

Return values:

• bull_reg — price makes a new low, indicator does not (bottom reversal signal)
• bear_reg — price makes a new high, indicator does not (top reversal signal)
• bull_hid — price makes a higher low, indicator a lower low (pullback within uptrend)
• bear_hid — price makes a lower high, indicator a higher high (bounce within downtrend)

BTC cycle indicators (BTC-USDT only)

All 5 indicators are based on price + time formulas; no on-chain data required:

Data requirements

Typical scenarios

"How does BTC 4H look technically?"

htx-cli spot-market kline -p symbol=btcusdt -p period=4hour -p size=200 | jq '.data' > btc.json
python scripts/indicators.py rsi --kline btc.json
python scripts/indicators.py macd --kline btc.json
python scripts/indicators.py supertrend --kline btc.json
python scripts/patterns.py scan --kline btc.json
python scripts/indicators.py divergence --kline btc.json
# AI synthesizes all outputs to make a judgment

"Is ETH overbought?"

python scripts/indicators.py rsi --kline eth4h.json
# rsi > 70 means overbought

"BTC long-term valuation right now"

htx-cli spot-market kline -p symbol=btcusdt -p period=1day -p size=400 | jq '.data' > btc1d.json
python scripts/cycle.py all --kline btc1d.json

Relationship with other skills

• Data source: depends on htx/spot-market or htx/futures-market for klines
• Upper-layer orchestration: invoked by htx/ta-master as the "price/volume" pillar

Installation

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