HTX Derivatives Analyst
Layer 2 analytical skill that orchestrates 5 atomic Layer 1 skills into a unified pressure score. Use when the user wants a one-shot read on whether a perpetual is overheated, where the squeeze risk lies, and what the directional bias is.
When to use this skill
- "How crowded is BTC perpetual right now?"
- "Squeeze risk on ETH-USDT?"
- "Are longs or shorts in trouble on SOL?"
- "Should I open a futures position on BTC right now?"
- "Give me a derivatives pressure read on ETH"
- "Why is the perpetual moving so fast?"
For pure technical analysis on price, prefer htx-technical-analysis. For just one signal (e.g. only funding), prefer the focused Layer 1 skill.
Underlying tools
This skill does not call REST endpoints directly. It composes Layer 1 skills:
| Layer 1 skill | What this analyst pulls from it |
|---|---|
https://github.com/htx-exchange/htx-skills-hub/funding-rate |
current rate + 30-period history |
https://github.com/htx-exchange/htx-skills-hub/oi-tracker |
current OI + 24h trend |
https://github.com/htx-exchange/htx-skills-hub/elite-positioning |
account ratio + position ratio |
https://github.com/htx-exchange/htx-skills-hub/liquidation-stream |
recent 7d liquidations |
https://github.com/htx-exchange/htx-skills-hub/mark-price |
basis kline (last 20 bars) |
If those skills aren't installed, install them first:
htx-cli skill install funding-rate
htx-cli skill install oi-tracker
htx-cli skill install elite-positioning
htx-cli skill install liquidation-stream
htx-cli skill install mark-price
Standard workflow
For a contract <code> (e.g. BTC-USDT), execute these in parallel where possible:
# 1. Funding rate
htx-cli futures market funding-rate <code> --json
htx-cli futures market historical-funding-rate --contract-code <code> --json
# 2. OI snapshot + history
htx-cli futures call GET /linear-swap-api/v1/swap_open_interest -p contract_code=<code> --json
htx-cli futures call GET /linear-swap-ex/market/his_open_interest \
-p contract_code=<code> -p period=4hour -p size=12 --json
# 3. Elite L/S ratio (both versions)
htx-cli futures call GET /linear-swap-api/v1/swap_elite_account_ratio \
-p contract_code=<code> -p period=1hour --json
htx-cli futures call GET /linear-swap-api/v1/swap_elite_position_ratio \
-p contract_code=<code> -p period=1hour --json
# 4. Recent liquidations (7d)
htx-cli futures market liquidation-orders <code> --json
# 5. Basis kline (recent)
htx-cli futures call GET /index/market/history/linear_swap_basis \
-p contract_code=<code> -p period=60min -p basis_price_type=close -p size=24 --json
Composite pressure score
Score each dimension on 0-100 (higher = more crowded / overheated), then weighted-average:
| Dimension | Weight | Computation |
|---|---|---|
| Funding | 25% | percentile of current rate vs last 30 periods. >85 pct = score 90+ |
| OI surge | 20% | 24h OI Δ%. ≥+15% → score 90; flat → 50; ≥-15% → score 10 |
| Elite divergence | 20% | abs(account_ratio − position_ratio) / account_ratio. >0.3 = score 80+ |
| Liquidation cluster | 15% | total 24h liq value / 30d avg. >2× → score 85+ |
| Basis stretch | 20% | percentile of current basis vs last 24h. extreme tail = high score |
Composite 0-100 → label:
| Score | Label | Interpretation |
|---|---|---|
| 0-30 | low | Calm; positions may unwind quietly |
| 31-55 | balanced | Healthy two-sided market |
| 56-75 | crowded | One side is concentrated; reversal risk rising |
| 76-100 | extreme | High-probability cleanout incoming |
Squeeze risk classification
Independent of the overall score, also flag squeeze direction:
| Setup | Verdict |
|---|---|
| Funding > 90 pct + elite_account_ratio > 1.5 + recent long-liq surge | long_squeeze (price likely capitulates lower) |
| Funding < 10 pct + elite_account_ratio < 0.7 + recent short-liq surge | short_squeeze (price likely rips higher) |
| Mixed | none |
Output structure
{
"skill": "derivatives-analyst",
"symbol": "BTC-USDT",
"timestamp": "2026-...",
"summary": {
"market_state": "overheated_long | overheated_short | balanced | deleveraging",
"leverage_risk": "high | medium | low",
"squeeze_risk": "long_squeeze | short_squeeze | none",
"signal_strength": 0-100,
"one_liner": "BTC perp funding 0.045% (95th pct), OI +12% 24h, elite position ratio 0.55 — heavy short capital, squeeze risk rising"
},
"components": {
"funding": {"current": 0.00045, "percentile_30p": 95, "score": 92},
"oi": {"current": 12_500_000_000, "delta_24h_pct": 12, "trend": "rising", "score": 75},
"elite": {"account_ratio": 1.85, "position_ratio": 0.55, "divergence": 0.7, "score": 88},
"liquidations": {"total_24h_usd": 45_000_000, "vs_30d_avg": 2.4, "long_pct": 70, "score": 82},
"basis": {"current_pct": 0.18, "percentile_24h": 92, "score": 78},
"composite_score": 82
},
"actionable": {
"suggested_action": "avoid new long positions / consider partial profit-taking on existing longs",
"trigger_conditions": "if OI keeps climbing without price response, expect violent unwind"
},
"risk_warning": "Past pressure regimes have ~65% reversal probability within 48h once score > 80."
}
What this skill explicitly does NOT do
- ⚠️ No all-market long/short ratio — HTX only exposes "elite" (top trader) ratios. Retail-vs-elite divergence is not directly observable.
- ⚠️ No Taker buy/sell volume — HTX has no dedicated endpoint; would need to derive from trade stream.
- ⚠️ No liquidation heatmap — no on-platform data; would require external CoinGlass integration.
These data gaps are documented so the agent can tell the user "we have X confidence" rather than over-claiming.
Related skills
https://github.com/htx-exchange/htx-skills-hub/funding-rate,https://github.com/htx-exchange/htx-skills-hub/oi-tracker,https://github.com/htx-exchange/htx-skills-hub/elite-positioning,https://github.com/htx-exchange/htx-skills-hub/liquidation-stream,https://github.com/htx-exchange/htx-skills-hub/mark-price— data sourceshttps://github.com/htx-exchange/htx-skills-hub/technical-analysis— pair derivatives pressure with price-action readhttps://github.com/htx-exchange/htx-skills-hub/sentiment-analyst— pair derivatives crowdedness with broader sentiment