Risk History — Historical Streak Analytics

The /risk/history endpoint returns historical streak analytics — the best and worst streaks ever, broken down globally, per symbol (BTC, ETH, SOL, XRP, etc.) and per timeframe (5m, 15m, 1h, 4h, 1d), with date ranges for each streak. This is separate from the live /risk endpoint so that expensive all-history aggregation does not affect live decision-making. Only counts decisions from the default preset (standard agent pipeline, not playground experiments).

Endpoint

GET /api/feeds/kronos/risk/history

Parameters

This endpoint takes no parameters — it returns all historical streak data.

Pricing

$0.02 per request (payment after delivery — settled only if the response is successfully sent).

Example request

curl -H "X-PAYMENT: <x402-payment-header>" \
  "https://kronos.seshat.markets/api/feeds/kronos/risk/history"

Example response (truncated)

{
  "global": {
    "best_streak": {
      "count": 14,
      "start_date": "2025-03-12",
      "end_date": "2025-03-26"
    },
    "worst_streak": {
      "count": -7,
      "start_date": "2025-01-08",
      "end_date": "2025-01-15"
    }
  },
  "by_symbol": {
    "btc_usdt": {
      "best_streak": { "count": 11, "start_date": "2025-03-12", "end_date": "2025-03-23" },
      "worst_streak": { "count": -5, "start_date": "2025-01-08", "end_date": "2025-01-13" }
    },
    "eth_usdt": {
      "best_streak": { "count": 9, "start_date": "2025-04-01", "end_date": "2025-04-10" },
      "worst_streak": { "count": -6, "start_date": "2025-02-14", "end_date": "2025-02-20" }
    }
  },
  "by_timeframe": {
    "1h": {
      "best_streak": { "count": 8, "start_date": "2025-03-15", "end_date": "2025-03-23" },
      "worst_streak": { "count": -4, "start_date": "2025-01-10", "end_date": "2025-01-14" }
    },
    "4h": {
      "best_streak": { "count": 6, "start_date": "2025-04-02", "end_date": "2025-04-08" },
      "worst_streak": { "count": -3, "start_date": "2025-02-16", "end_date": "2025-02-19" }
    }
  }
}

Use cases

Confidence Calibration

If the model is on a 14-win streak, you can trust the next prediction more. If it just came off a 7-loss streak, dial down position size or wait for recovery.

Regime Detection

Long winning streaks correlate with trending markets; losing streaks with choppy/ranging markets. Use streak data to infer the current regime without external indicators.

Per-Symbol Trust Profiles

BTC might have a best streak of 11 while ETH's is 9 — BTC signals are more reliable in trending conditions. Build per-symbol trust profiles from historical streaks.

Timeframe Filtering

If the 1h timeframe has longer winning streaks than 5m, filter out 5m signals from your strategy. Streak length is a proxy for timeframe reliability.

Frequently asked questions

What is a crypto prediction streak analytics API?
It returns the best and worst prediction streaks ever — globally, per symbol (BTC, ETH, SOL), and per timeframe (5m, 15m, 1h, 4h, 1d) with date ranges. Use it to calibrate confidence based on whether the model is on a winning or losing streak.
How do I know if a crypto prediction model is hot or cold?
Call GET /api/feeds/kronos/risk/history. If the model is on a 14-win streak, trust the next prediction more. If it just came off a 7-loss streak, dial down position size or wait for recovery.
How much does the risk history API cost?
$0.02 per request (payment after delivery). Pay with USDC on Solana or Base via x402. Check the live risk state for free at GET /api/feeds/kronos/risk.
Can I use streak data for crypto regime detection?
Yes. Long winning streaks correlate with trending markets; losing streaks with choppy/ranging markets. Use streak data to infer the current regime without external indicators, then adjust exposure accordingly.

Related endpoints