Cerbero-mcp/src/cerbero_mcp/exchanges/deribit/client.py

from __future__ import annotations

import contextlib
import time
from dataclasses import dataclass, field
from typing import Any

from cerbero_mcp.common import indicators as ind
from cerbero_mcp.common import microstructure as micro
from cerbero_mcp.common import options as opt
from cerbero_mcp.common.http import async_client

BASE_LIVE = "https://www.deribit.com/api/v2"
BASE_TESTNET = "https://test.deribit.com/api/v2"

RESOLUTION_MAP = {
    "1m": "1",
    "5m": "5",
    "15m": "15",
    "1h": "60",
    "4h": "240",
    "1d": "1D",
}


@dataclass
class DeribitClient:
    client_id: str
    client_secret: str
    testnet: bool = True
    base_url_override: str | None = None
    _token: str | None = field(default=None, init=False, repr=False)
    _token_expires_at: float = field(default=0.0, init=False, repr=False)

    @property
    def base_url(self) -> str:
        if self.base_url_override:
            return self.base_url_override
        return BASE_TESTNET if self.testnet else BASE_LIVE

    # ── Auth ─────────────────────────────────────────────────────

    async def _authenticate(self) -> str:
        url = f"{self.base_url}/public/auth"
        params = {
            "grant_type": "client_credentials",
            "client_id": self.client_id,
            "client_secret": self.client_secret,
        }
        async with async_client(timeout=15.0) as http:
            resp = await http.get(url, params=params)
            data = resp.json()
        result = data["result"]
        self._token = result["access_token"]
        self._token_expires_at = time.monotonic() + result.get("expires_in", 900) - 30
        return self._token

    async def _get_token(self) -> str:
        if self._token is None or time.monotonic() >= self._token_expires_at:
            await self._authenticate()
        return self._token  # type: ignore[return-value]

    async def _request(self, method: str, params: dict[str, Any] | None = None) -> dict:
        is_private = method.startswith("private/")
        if is_private:
            await self._get_token()

        url = f"{self.base_url}/{method}"
        request_params = dict(params) if params else {}
        headers: dict[str, str] = {}
        if is_private and self._token:
            headers["Authorization"] = f"Bearer {self._token}"

        async with async_client(timeout=15.0) as http:
            resp = await http.get(url, params=request_params, headers=headers)
            data = resp.json()

            if "result" not in data:
                error = data.get("error", {})
                error_code = error.get("code", 0) if isinstance(error, dict) else 0
                error_msg = error.get("message", str(data)) if isinstance(error, dict) else str(error)
                # Re-authenticate on auth errors and retry once
                if is_private and error_code in (13004, 13009, 13015):
                    self._token = None
                    await self._authenticate()
                    headers["Authorization"] = f"Bearer {self._token}"
                    resp = await http.get(url, params=request_params, headers=headers)
                    data = resp.json()
                    if "result" in data:
                        return data  # type: ignore[no-any-return]
                return {"result": None, "error": error_msg}

        return data  # type: ignore[no-any-return]

    # ── Read tools ───────────────────────────────────────────────

    def is_testnet(self) -> dict:
        return {"testnet": self.testnet, "base_url": self.base_url}

    async def health(self) -> dict:
        """Probe minimo per /health/ready: nessuna chiamata di rete."""
        return {"status": "ok", "testnet": self.testnet}

    async def get_ticker(self, instrument_name: str) -> dict:
        import datetime as _dt
        raw = await self._request("public/ticker", {"instrument_name": instrument_name})
        r = raw.get("result") or {}
        is_perp = instrument_name.upper().endswith("-PERPETUAL")
        greeks = r.get("greeks")
        if is_perp and greeks is None:
            greeks = {"delta": 1.0, "gamma": 0.0, "vega": 0.0, "theta": 0.0, "rho": 0.0}
        return {
            "instrument_name": instrument_name,
            "last_price": r.get("last_price"),
            "mark_price": r.get("mark_price"),
            "bid": r.get("best_bid_price"),
            "ask": r.get("best_ask_price"),
            "volume_24h": r.get("stats", {}).get("volume"),
            "open_interest": r.get("open_interest"),
            "greeks": greeks,
            "mark_iv": r.get("mark_iv"),
            "testnet": self.testnet,
            "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat(),
        }

    async def get_ticker_batch(self, instrument_names: list[str]) -> dict:
        """Fetch multiple tickers in parallel (max 20)."""
        import asyncio
        import datetime as _dt

        if not instrument_names:
            return {"tickers": [], "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat()}
        if len(instrument_names) > 20:
            return {"error": "max 20 instruments per batch"}

        results = await asyncio.gather(
            *[self.get_ticker(n) for n in instrument_names],
            return_exceptions=True,
        )
        tickers = []
        errors = []
        for name, res in zip(instrument_names, results, strict=True):
            if isinstance(res, Exception):
                errors.append({"instrument": name, "error": str(res)})
            else:
                tickers.append(res)
        return {
            "tickers": tickers,
            "errors": errors,
            "count": len(tickers),
            "testnet": self.testnet,
            "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat(),
        }

    async def get_instruments(
        self,
        currency: str,
        kind: str | None = None,
        expired: bool = False,
        expiry_from: str | None = None,
        expiry_to: str | None = None,
        strike_min: float | None = None,
        strike_max: float | None = None,
        min_open_interest: float | None = None,
        limit: int = 100,
        offset: int = 0,
    ) -> dict:
        import asyncio
        import datetime as _dt

        def _parse_iso(s: str | None) -> int | None:
            if not s:
                return None
            try:
                dt = _dt.datetime.fromisoformat(s)
            except ValueError:
                dt = _dt.datetime.strptime(s, "%Y-%m-%d")
            return int(dt.timestamp() * 1000)

        expiry_from_ms = _parse_iso(expiry_from)
        expiry_to_ms = _parse_iso(expiry_to)

        params: dict[str, Any] = {"currency": currency, "expired": expired}
        if kind:
            params["kind"] = kind

        # CER-008: fetch metadata + book_summary in parallelo per popolare OI.
        # `public/get_instruments` non include open_interest; book_summary sì.
        summary_params: dict[str, Any] = {"currency": currency}
        if kind:
            summary_params["kind"] = kind
        instruments_raw, summary_raw = await asyncio.gather(
            self._request("public/get_instruments", params),
            self._request("public/get_book_summary_by_currency", summary_params),
            return_exceptions=True,
        )
        raw = instruments_raw if isinstance(instruments_raw, dict) else {}  # type: ignore[has-type]
        summary_items = (
            summary_raw.get("result") if isinstance(summary_raw, dict) else None  # type: ignore[has-type]
        ) or []
        oi_by_name: dict[str, float] = {}
        for s in summary_items:
            name = s.get("instrument_name")
            oi = s.get("open_interest")
            if name and oi is not None:
                with contextlib.suppress(TypeError, ValueError):
                    oi_by_name[name] = float(oi)

        all_items = raw.get("result") or []
        filtered: list[dict] = []
        for i in all_items:
            exp_ms = i.get("expiration_timestamp")
            strike = i.get("strike")
            name = i.get("instrument_name")
            # CER-008: popola OI dal book_summary se mancante
            oi = i.get("open_interest")
            if oi is None and name in oi_by_name:
                oi = oi_by_name[name]
                i["open_interest"] = oi
            if expiry_from_ms is not None and exp_ms is not None and exp_ms < expiry_from_ms:
                continue
            if expiry_to_ms is not None and exp_ms is not None and exp_ms > expiry_to_ms:
                continue
            if strike_min is not None and strike is not None and strike < strike_min:
                continue
            if strike_max is not None and strike is not None and strike > strike_max:
                continue
            if (
                min_open_interest is not None
                and oi is not None
                and oi < min_open_interest
            ):
                continue
            filtered.append(i)

        total = len(filtered)
        page = filtered[offset : offset + limit]
        instruments = [
            {
                "name": i.get("instrument_name"),
                "strike": i.get("strike"),
                "expiry": i.get("expiration_timestamp"),
                "option_type": i.get("option_type"),
                "tick_size": i.get("tick_size"),
                "min_trade_amount": i.get("min_trade_amount"),
                "open_interest": i.get("open_interest"),
            }
            for i in page
        ]
        return {
            "instruments": instruments,
            "total": total,
            "offset": offset,
            "limit": limit,
            "has_more": offset + limit < total,
            "testnet": self.testnet,
        }

    async def get_orderbook(self, instrument_name: str, depth: int = 10) -> dict:
        import datetime as _dt
        raw = await self._request(
            "public/get_order_book", {"instrument_name": instrument_name, "depth": depth}
        )
        r = raw.get("result") or {}
        return {
            "bids": r.get("bids", []),
            "asks": r.get("asks", []),
            "timestamp": r.get("timestamp"),
            "testnet": self.testnet,
            "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat(),
        }

    async def get_orderbook_imbalance(self, instrument_name: str, depth: int = 10) -> dict:
        """Microstructure: bid/ask imbalance + microprice + slope su top-N livelli."""
        ob = await self.get_orderbook(instrument_name, depth=max(depth, 10))
        result = micro.orderbook_imbalance(ob.get("bids") or [], ob.get("asks") or [], depth=depth)
        return {
            "instrument_name": instrument_name,
            "depth": depth,
            **result,
            "timestamp": ob.get("timestamp"),
            "testnet": self.testnet,
        }

    async def get_positions(self, currency: str = "USDC") -> list:
        raw = await self._request("private/get_positions", {"currency": currency})
        result = raw.get("result") or []
        positions = []
        for p in result:
            size = p.get("size", 0)
            if size == 0:
                continue
            positions.append({
                "instrument": p.get("instrument_name"),
                "size": abs(size),
                "direction": "long" if size > 0 else "short",
                "avg_price": p.get("average_price"),
                "mark_price": p.get("mark_price"),
                "unrealized_pnl": p.get("floating_profit_loss"),
                "realized_pnl": p.get("realized_profit_loss"),
                "leverage": p.get("leverage"),
            })
        return positions

    async def get_account_summary(self, currency: str = "USDC") -> dict:
        raw = await self._request("private/get_account_summary", {"currency": currency})
        r = raw.get("result")
        if not r:
            return {
                "equity": 0,
                "balance": 0,
                "margin_balance": 0,
                "available_funds": 0,
                "unrealized_pnl": 0,
                "total_pnl": 0,
                "testnet": self.testnet,
                "error": raw.get("error", "no result"),
            }
        return {
            "equity": r.get("equity", 0),
            "balance": r.get("balance", 0),
            "margin_balance": r.get("margin_balance", 0),
            "available_funds": r.get("available_funds", 0),
            "unrealized_pnl": r.get("unrealized_pnl", 0),
            "total_pnl": r.get("total_pnl", 0),
            "testnet": self.testnet,
        }

    async def get_trade_history(
        self, limit: int = 100, instrument_name: str | None = None
    ) -> list:
        if instrument_name:
            raw = await self._request(
                "private/get_user_trades_by_instrument",
                {"instrument_name": instrument_name, "count": limit},
            )
        else:
            raw = await self._request(
                "private/get_user_trades_by_currency",
                {"currency": "BTC", "count": limit},
            )
        result = raw.get("result") or {}
        if not isinstance(result, dict):
            return []
        trades_raw = result.get("trades") or []
        trades = []
        for t in trades_raw:
            if not isinstance(t, dict):
                continue
            trades.append({
                "instrument": t.get("instrument_name"),
                "direction": t.get("direction"),
                "price": t.get("price"),
                "amount": t.get("amount"),
                "fee": t.get("fee"),
                "timestamp": t.get("timestamp"),
                "order_id": t.get("order_id"),
            })
        return trades

    async def get_historical(
        self, instrument: str, start_date: str, end_date: str, resolution: str
    ) -> dict:
        # Convert ISO date strings to millisecond timestamps
        import datetime as _dt

        def _to_ms(date_str: str) -> int:
            try:
                dt = _dt.datetime.fromisoformat(date_str)
            except ValueError:
                dt = _dt.datetime.strptime(date_str, "%Y-%m-%d")
            return int(dt.timestamp() * 1000)

        start_ms = _to_ms(start_date)
        end_ms = _to_ms(end_date)
        res = RESOLUTION_MAP.get(resolution, resolution)
        raw = await self._request(
            "public/get_tradingview_chart_data",
            {
                "instrument_name": instrument,
                "start_timestamp": start_ms,
                "end_timestamp": end_ms,
                "resolution": res,
            },
        )
        r = raw.get("result") or {}
        candles = []
        ticks = r.get("ticks", []) or []
        opens = r.get("open", []) or []
        highs = r.get("high", []) or []
        lows = r.get("low", []) or []
        closes = r.get("close", []) or []
        volumes = r.get("volume", []) or []
        for idx, ts in enumerate(ticks):
            if idx >= min(len(opens), len(highs), len(lows), len(closes), len(volumes)):
                break
            candles.append({
                "timestamp": ts,
                "open": opens[idx],
                "high": highs[idx],
                "low": lows[idx],
                "close": closes[idx],
                "volume": volumes[idx],
            })
        return {"candles": candles}

    async def get_dvol(
        self,
        currency: str,
        start_date: str,
        end_date: str,
        resolution: str = "1D",
    ) -> dict:
        import datetime as _dt

        def _to_ms(date_str: str) -> int:
            try:
                dt = _dt.datetime.fromisoformat(date_str)
            except ValueError:
                dt = _dt.datetime.strptime(date_str, "%Y-%m-%d")
            return int(dt.timestamp() * 1000)

        start_ms = _to_ms(start_date)
        end_ms = _to_ms(end_date)
        res = RESOLUTION_MAP.get(resolution, resolution)
        raw = await self._request(
            "public/get_volatility_index_data",
            {
                "currency": currency.upper(),
                "start_timestamp": start_ms,
                "end_timestamp": end_ms,
                "resolution": res,
            },
        )
        r = raw.get("result") or {}
        rows = r.get("data") or []
        candles = [
            {
                "timestamp": row[0],
                "open": row[1],
                "high": row[2],
                "low": row[3],
                "close": row[4],
            }
            for row in rows
            if len(row) >= 5
        ]
        latest = candles[-1]["close"] if candles else None
        return {"currency": currency.upper(), "latest": latest, "candles": candles}

    async def get_gex(
        self,
        currency: str,
        expiry_from: str | None = None,
        expiry_to: str | None = None,
        top_n_strikes: int = 50,
    ) -> dict:
        import asyncio
        import datetime as _dt

        currency = currency.upper()
        try:
            idx_tk = await self.get_ticker(f"{currency}-PERPETUAL")
            spot = float(idx_tk.get("mark_price") or 0)
        except Exception:
            spot = 0.0

        chain = await self.get_instruments(
            currency=currency,
            kind="option",
            expiry_from=expiry_from,
            expiry_to=expiry_to,
            limit=2000,
        )
        items = chain.get("instruments", [])
        items.sort(key=lambda x: -(x.get("open_interest") or 0))
        top = items[:top_n_strikes]

        async def _ticker(name: str) -> dict:
            try:
                return await self.get_ticker(name)
            except Exception:
                return {}

        tickers = await asyncio.gather(*[_ticker(i["name"]) for i in top])

        by_strike: dict[float, dict[str, float]] = {}
        for meta, tk in zip(top, tickers, strict=True):
            strike = meta.get("strike")
            if strike is None:
                continue
            greeks = tk.get("greeks") or {}
            gamma = greeks.get("gamma")
            oi = meta.get("open_interest") or 0
            if gamma is None or spot <= 0:
                continue
            gex_contribution = float(gamma) * oi * (spot ** 2) * 0.01
            entry = by_strike.setdefault(
                float(strike), {"strike": float(strike), "call_gex": 0.0, "put_gex": 0.0}
            )
            if meta.get("option_type") == "call":
                entry["call_gex"] += gex_contribution
            else:
                entry["put_gex"] -= gex_contribution

        rows = []
        for s in sorted(by_strike.keys()):
            e = by_strike[s]
            e["net_gex"] = e["call_gex"] + e["put_gex"]
            rows.append(e)

        total_gex = sum(r["net_gex"] for r in rows)
        zero_gamma = None
        for a, b in zip(rows, rows[1:], strict=False):
            if (a["net_gex"] < 0 <= b["net_gex"]) or (a["net_gex"] > 0 >= b["net_gex"]):
                denom = b["net_gex"] - a["net_gex"]
                if denom != 0:
                    frac = -a["net_gex"] / denom
                    zero_gamma = round(a["strike"] + frac * (b["strike"] - a["strike"]), 2)
                    break

        max_gex_level = max(rows, key=lambda r: r["net_gex"])["strike"] if rows else None
        min_gex_level = min(rows, key=lambda r: r["net_gex"])["strike"] if rows else None

        return {
            "currency": currency,
            "expiry_from": expiry_from,
            "expiry_to": expiry_to,
            "spot_price": spot,
            "total_gex_usd": round(total_gex, 2),
            "zero_gamma_level": zero_gamma,
            "gex_by_strike": [
                {
                    "strike": r["strike"],
                    "call_gex": round(r["call_gex"], 2),
                    "put_gex": round(r["put_gex"], 2),
                    "net_gex": round(r["net_gex"], 2),
                }
                for r in rows
            ],
            "max_gex_level": max_gex_level,
            "min_gex_level": min_gex_level,
            "strikes_analyzed": len(rows),
            "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat(),
            "testnet": self.testnet,
        }

    async def _fetch_chain_legs(
        self,
        currency: str,
        expiry_from: str | None = None,
        expiry_to: str | None = None,
        top_n_strikes: int = 50,
    ) -> tuple[float, list[dict[str, Any]]]:
        """Fetch chain options + ticker per top-N strikes per OI; restituisce
        (spot, legs[]) con campi normalizzati per le funzioni in cerbero_mcp.common.options.
        """
        import asyncio

        currency = currency.upper()
        try:
            idx_tk = await self.get_ticker(f"{currency}-PERPETUAL")
            spot = float(idx_tk.get("mark_price") or 0)
        except Exception:
            spot = 0.0

        chain = await self.get_instruments(
            currency=currency,
            kind="option",
            expiry_from=expiry_from,
            expiry_to=expiry_to,
            limit=2000,
        )
        items = chain.get("instruments", [])
        items.sort(key=lambda x: -(x.get("open_interest") or 0))
        top = items[:top_n_strikes]

        async def _ticker(name: str) -> dict:
            try:
                return await self.get_ticker(name)
            except Exception:
                return {}

        tickers = await asyncio.gather(*[_ticker(i["name"]) for i in top])
        legs: list[dict[str, Any]] = []
        for meta, tk in zip(top, tickers, strict=True):
            greeks = tk.get("greeks") or {}
            legs.append({
                "strike": meta.get("strike"),
                "option_type": meta.get("option_type"),
                "oi": meta.get("open_interest") or 0,
                "iv": tk.get("mark_iv"),
                "delta": greeks.get("delta"),
                "gamma": greeks.get("gamma"),
                "vanna": greeks.get("vanna"),
                "charm": greeks.get("charm"),
                "vega": greeks.get("vega"),
            })
        return spot, legs

    async def get_dealer_gamma_profile(
        self,
        currency: str,
        expiry_from: str | None = None,
        expiry_to: str | None = None,
        top_n_strikes: int = 50,
    ) -> dict:
        """Net dealer gamma per strike (assume dealer short calls/long puts).
        Identifica il gamma flip level: sopra → mercato pinning, sotto → squeeze.
        """
        import datetime as _dt
        spot, legs = await self._fetch_chain_legs(currency, expiry_from, expiry_to, top_n_strikes)
        result = opt.dealer_gamma_profile(legs, spot)
        return {
            "currency": currency.upper(),
            "spot_price": spot,
            **result,
            "strikes_analyzed": len(result["by_strike"]),
            "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat(),
            "testnet": self.testnet,
        }

    async def get_vanna_charm(
        self,
        currency: str,
        expiry_from: str | None = None,
        expiry_to: str | None = None,
        top_n_strikes: int = 50,
    ) -> dict:
        """Vanna (∂delta/∂IV) e Charm (∂delta/∂t) aggregati pesati per OI.
        Vanna positiva: dealer compra spot quando IV sale.
        Charm negativa: time decay erode delta hedging.
        """
        import datetime as _dt
        spot, legs = await self._fetch_chain_legs(currency, expiry_from, expiry_to, top_n_strikes)
        result = opt.vanna_charm_aggregate(legs, spot)
        return {
            "currency": currency.upper(),
            **result,
            "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat(),
            "testnet": self.testnet,
        }

    async def get_oi_weighted_skew(
        self,
        currency: str,
        expiry_from: str | None = None,
        expiry_to: str | None = None,
        top_n_strikes: int = 100,
    ) -> dict:
        """Skew aggregato pesato OI: IV media puts - calls. Positivo = paura.
        """
        import datetime as _dt
        _, legs = await self._fetch_chain_legs(currency, expiry_from, expiry_to, top_n_strikes)
        result = opt.oi_weighted_skew(legs)
        return {
            "currency": currency.upper(),
            **result,
            "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat(),
            "testnet": self.testnet,
        }

    async def get_smile_asymmetry(
        self,
        currency: str,
        expiry_from: str | None = None,
        expiry_to: str | None = None,
        top_n_strikes: int = 100,
    ) -> dict:
        """Asymmetry IV otm-puts vs otm-calls. Positivo = put-side richer."""
        import datetime as _dt
        spot, legs = await self._fetch_chain_legs(currency, expiry_from, expiry_to, top_n_strikes)
        result = opt.smile_asymmetry(legs, spot)
        return {
            "currency": currency.upper(),
            "spot_price": spot,
            **result,
            "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat(),
            "testnet": self.testnet,
        }

    async def get_atm_vs_wings_vol(
        self,
        currency: str,
        expiry_from: str | None = None,
        expiry_to: str | None = None,
        top_n_strikes: int = 100,
    ) -> dict:
        """IV ATM vs IV alle ali 25-delta. wing_richness > 0 → smile (kurtosis)."""
        import datetime as _dt
        spot, legs = await self._fetch_chain_legs(currency, expiry_from, expiry_to, top_n_strikes)
        result = opt.atm_vs_wings_vol(legs, spot)
        return {
            "currency": currency.upper(),
            "spot_price": spot,
            **result,
            "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat(),
            "testnet": self.testnet,
        }

    async def get_pc_ratio(self, currency: str) -> dict:
        import datetime as _dt

        currency = currency.upper()
        raw = await self._request(
            "public/get_book_summary_by_currency",
            {"currency": currency, "kind": "option"},
        )
        rows = raw.get("result") or []
        call_oi = 0.0
        put_oi = 0.0
        call_vol = 0.0
        put_vol = 0.0
        for r in rows:
            name = r.get("instrument_name", "")
            oi = r.get("open_interest") or 0
            vol = r.get("volume") or 0
            if name.endswith("-C"):
                call_oi += oi
                call_vol += vol
            elif name.endswith("-P"):
                put_oi += oi
                put_vol += vol

        def _ratio(put: float, call: float) -> float | None:
            return round(put / call, 4) if call > 0 else None

        def _interp(ratio: float | None) -> str:
            if ratio is None:
                return "insufficient_data"
            if ratio > 1.0:
                return "puts_dominant"
            if ratio < 0.7:
                return "calls_dominant"
            return "balanced"

        pc_oi = _ratio(put_oi, call_oi)
        pc_vol = _ratio(put_vol, call_vol)

        return {
            "currency": currency,
            "pc_ratio_oi": pc_oi,
            "pc_ratio_volume_24h": pc_vol,
            "total_call_oi": call_oi,
            "total_put_oi": put_oi,
            "total_call_volume_24h": call_vol,
            "total_put_volume_24h": put_vol,
            "interpretation": {
                "oi": _interp(pc_oi),
                "volume_24h": _interp(pc_vol),
                "percentile_90d": None,
            },
            "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat(),
            "testnet": self.testnet,
        }

    async def get_skew_25d(self, currency: str, expiry: str) -> dict:
        import asyncio
        import datetime as _dt

        currency = currency.upper()
        put_task = self.find_by_delta(currency, expiry, -0.25, "put", 1, 0, 0)
        call_task = self.find_by_delta(currency, expiry, 0.25, "call", 1, 0, 0)
        try:
            idx_tk = await self.get_ticker(f"{currency}-PERPETUAL")
            spot = float(idx_tk.get("mark_price") or 0)
        except Exception:
            spot = 0.0
        put_res, call_res = await asyncio.gather(put_task, call_task)

        put_best = put_res.get("best_match") or {}
        call_best = call_res.get("best_match") or {}
        put_iv = put_best.get("mark_iv")
        call_iv = call_best.get("mark_iv")

        # ATM IV: find instrument closest to spot on this expiry
        exp_dt = _dt.datetime.fromisoformat(expiry) if "T" in expiry or "-" in expiry else _dt.datetime.strptime(expiry, "%Y-%m-%d")
        chain = await self.get_instruments(
            currency=currency,
            kind="option",
            expiry_from=expiry,
            expiry_to=(exp_dt + _dt.timedelta(days=1)).strftime("%Y-%m-%d"),
            limit=500,
        )
        items = chain.get("instruments", [])
        atm_iv = None
        if items and spot > 0:
            call_items = [i for i in items if i.get("option_type") == "call"]
            if call_items:
                call_items.sort(key=lambda x: abs((x.get("strike") or 0) - spot))
                atm_tk = await self.get_ticker(call_items[0]["name"])
                atm_iv = atm_tk.get("mark_iv")

        skew = None
        if put_iv is not None and call_iv is not None:
            skew = round(put_iv - call_iv, 4)

        if skew is None:
            skew_sign = "unknown"
        elif skew > 1:
            skew_sign = "puts_rich"
        elif skew < -1:
            skew_sign = "calls_rich"
        else:
            skew_sign = "neutral"

        butterfly = None
        if put_iv is not None and call_iv is not None and atm_iv is not None:
            butterfly = round((put_iv + call_iv) / 2 - atm_iv, 4)

        return {
            "currency": currency,
            "expiry": expiry,
            "put_25d_iv": put_iv,
            "call_25d_iv": call_iv,
            "atm_iv": atm_iv,
            "skew_25d": skew,
            "skew_sign": skew_sign,
            "risk_reversal_25d": round(-skew, 4) if skew is not None else None,
            "butterfly_25d": butterfly,
            "skew_percentile_90d": None,
            "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat(),
            "testnet": self.testnet,
        }

    async def get_term_structure(self, currency: str) -> dict:
        import asyncio
        import datetime as _dt

        currency = currency.upper()
        try:
            idx_tk = await self.get_ticker(f"{currency}-PERPETUAL")
            spot = float(idx_tk.get("mark_price") or 0)
        except Exception:
            spot = 0.0

        chain = await self.get_instruments(
            currency=currency, kind="option", limit=2000, offset=0
        )
        items = chain.get("instruments", [])
        now_ms = int(_dt.datetime.now(_dt.UTC).timestamp() * 1000)
        by_exp: dict[int, list[dict]] = {}
        for i in items:
            exp = i.get("expiry")
            if exp is None or exp < now_ms:
                continue
            by_exp.setdefault(int(exp), []).append(i)

        atm_names: list[tuple[int, str]] = []
        for exp_ms, ins in sorted(by_exp.items()):
            if spot > 0:
                ins.sort(key=lambda x: abs((x.get("strike") or 0) - spot))
            atm_names.append((exp_ms, ins[0]["name"]))

        async def _ticker(name: str) -> dict:
            try:
                return await self.get_ticker(name)
            except Exception:
                return {}

        tickers = await asyncio.gather(*[_ticker(n) for _, n in atm_names])

        ts = []
        for (exp_ms, name), tk in zip(atm_names, tickers, strict=True):
            iv = tk.get("mark_iv")
            if iv is None:
                continue
            exp_dt = _dt.datetime.fromtimestamp(exp_ms / 1000, _dt.UTC)
            dte = max(0, (exp_dt - _dt.datetime.now(_dt.UTC)).days)
            ts.append({
                "expiry": exp_dt.strftime("%Y-%m-%d"),
                "dte": dte,
                "atm_iv": iv,
                "atm_instrument": name,
            })

        shape = "flat"
        contango_steep = False
        calendar_opp = False
        if len(ts) >= 2:
            diffs = [ts[i + 1]["atm_iv"] - ts[i]["atm_iv"] for i in range(len(ts) - 1)]
            up = sum(1 for d in diffs if d > 0)
            down = sum(1 for d in diffs if d < 0)
            if up > len(diffs) / 2:
                shape = "contango"
            elif down > len(diffs) / 2:
                shape = "backwardation"
            short_term = next((x for x in ts if 8 <= x["dte"] <= 14), None)
            mid_term = next((x for x in ts if 35 <= x["dte"] <= 45), None)
            if short_term and mid_term and mid_term["atm_iv"] - short_term["atm_iv"] > 5:
                contango_steep = True
                calendar_opp = True

        return {
            "currency": currency,
            "spot": spot,
            "term_structure": ts,
            "shape": shape,
            "contango_steep": contango_steep,
            "calendar_spread_opportunity": calendar_opp,
            "testnet": self.testnet,
        }

    async def run_backtest(
        self,
        strategy_name: str,
        underlying: str = "BTC",
        lookback_days: int = 30,
        resolution: str = "4h",
        entry_rules: dict | None = None,
        exit_rules: dict | None = None,
    ) -> dict:
        """Heuristic backtest: OHLCV lookback + simple rule-based entry/exit.
        Approximate; does not simulate full options chain.
        """
        import datetime as _dt

        end = _dt.datetime.now(_dt.UTC)
        start = end - _dt.timedelta(days=lookback_days)
        historical = await self.get_historical(
            f"{underlying.upper()}-PERPETUAL",
            start.strftime("%Y-%m-%d"),
            end.strftime("%Y-%m-%d"),
            resolution,
        )
        candles = historical.get("candles", [])
        if len(candles) < 20:
            return {
                "strategy_name": strategy_name,
                "error": "insufficient history",
                "trades_simulated": 0,
                "recommendation": "reject",
            }

        closes = [c["close"] for c in candles]
        from cerbero_mcp.common import indicators as _ind

        rsi_thr_low = (entry_rules or {}).get("rsi_below", 35)
        rsi_thr_high = (entry_rules or {}).get("rsi_above", 65)
        target_pct = (exit_rules or {}).get("target_pct", 2.0)
        stop_pct = (exit_rules or {}).get("stop_pct", -1.5)
        max_hold_bars = (exit_rules or {}).get("max_hold_bars", 20)

        trades: list[dict] = []
        i = 14
        while i < len(closes) - 1:
            sub = closes[: i + 1]
            rsi_val = _ind.rsi(sub)
            if rsi_val is None:
                i += 1
                continue
            side: str | None = None
            if rsi_val < rsi_thr_low:
                side = "long"
            elif rsi_val > rsi_thr_high:
                side = "short"
            if side is None:
                i += 1
                continue
            entry = closes[i]
            exit_bar = None
            exit_price = entry
            for j in range(1, min(max_hold_bars, len(closes) - i)):
                p = closes[i + j]
                pct = (p - entry) / entry * 100 * (1 if side == "long" else -1)
                if pct >= target_pct or pct <= stop_pct:
                    exit_bar = j
                    exit_price = p
                    break
            if exit_bar is None:
                exit_bar = min(max_hold_bars, len(closes) - i - 1)
                exit_price = closes[i + exit_bar]
            pnl_pct = (exit_price - entry) / entry * 100 * (1 if side == "long" else -1)
            trades.append({
                "entry_idx": i,
                "side": side,
                "entry": entry,
                "exit": exit_price,
                "bars_held": exit_bar,
                "pnl_pct": pnl_pct,
            })
            i += exit_bar + 1

        if not trades:
            return {
                "strategy_name": strategy_name,
                "trades_simulated": 0,
                "recommendation": "reject",
                "notes": "no entries triggered",
            }

        wins = [t for t in trades if t["pnl_pct"] > 0]
        losses = [t for t in trades if t["pnl_pct"] <= 0]
        win_rate = len(wins) / len(trades)
        total_profit = sum(t["pnl_pct"] for t in wins)
        total_loss = -sum(t["pnl_pct"] for t in losses)
        profit_factor = (total_profit / total_loss) if total_loss > 0 else float("inf")
        equity = 0.0
        peak = 0.0
        max_dd = 0.0
        for t in trades:
            equity += t["pnl_pct"]
            peak = max(peak, equity)
            dd = peak - equity
            max_dd = max(max_dd, dd)
        mean_r = sum(t["pnl_pct"] for t in trades) / len(trades)
        var_r = sum((t["pnl_pct"] - mean_r) ** 2 for t in trades) / len(trades)
        std_r = var_r ** 0.5
        sharpe = round(mean_r / std_r, 2) if std_r > 0 else None

        recommendation = "reject"
        if win_rate > 0.55 and max_dd < 4.0 and profit_factor > 1.5:
            recommendation = "accept"
        elif win_rate > 0.50 and max_dd < 5.0 and profit_factor > 1.35:
            recommendation = "marginal"

        return {
            "strategy_name": strategy_name,
            "underlying": underlying.upper(),
            "lookback_days": lookback_days,
            "resolution": resolution,
            "trades_simulated": len(trades),
            "win_rate": round(win_rate, 3),
            "max_drawdown_pct": round(max_dd, 2),
            "profit_factor": round(profit_factor, 2) if profit_factor != float("inf") else None,
            "sharpe": sharpe,
            "recommendation": recommendation,
            "notes": "heuristic RSI-based backtest — approximate, not full options simulation",
            "testnet": self.testnet,
            "data_timestamp": _dt.datetime.now(_dt.UTC).isoformat(),
        }

    async def calculate_spread_payoff(
        self,
        legs: list[dict],
        quote_currency: str = "USD",
    ) -> dict:
        import asyncio
        import re as _re

        if not legs or len(legs) > 4:
            return {"error": "legs must be 1..4"}

        async def _fetch(name: str) -> dict:
            try:
                return await self.get_ticker(name)
            except Exception as e:
                return {"error": str(e)}

        tickers = await asyncio.gather(*[_fetch(l["instrument_name"]) for l in legs])

        enriched: list[dict] = []
        spot_est: float | None = None
        for leg, tk in zip(legs, tickers, strict=True):
            name = leg["instrument_name"]
            match = _re.match(
                r"^([A-Z]+)-(\d{1,2}[A-Z]{3}\d{2})-(\d+(?:\.\d+)?)-(C|P)$", name
            )
            if not match:
                return {"error": f"invalid option name: {name}"}
            coin, exp_str, strike_s, opt_type = match.groups()
            strike = float(strike_s)
            action = leg.get("action", "long").lower()
            qty = float(leg.get("quantity", 1))
            sign = 1 if action == "long" else -1
            mark_price = tk.get("mark_price") or 0.0
            greeks = tk.get("greeks") or {}
            if spot_est is None and mark_price:
                spot_est = float(strike)
            enriched.append({
                "name": name,
                "coin": coin,
                "expiry": exp_str,
                "strike": strike,
                "type": opt_type,
                "action": action,
                "quantity": qty,
                "sign": sign,
                "mark_price": float(mark_price),
                "greeks": greeks,
            })

        strikes = [l["strike"] for l in enriched]
        spot = spot_est or (sum(strikes) / len(strikes)) if strikes else 0.0
        if enriched and enriched[0]["coin"]:
            try:
                idx_name = f"{enriched[0]['coin']}-PERPETUAL"
                idx_tk = await self.get_ticker(idx_name)
                if idx_tk.get("mark_price"):
                    spot = float(idx_tk["mark_price"])
            except Exception:
                pass

        net_premium = sum(l["sign"] * l["quantity"] * l["mark_price"] for l in enriched)
        credit = net_premium < 0

        greeks_net = {k: 0.0 for k in ("delta", "gamma", "vega", "theta", "rho")}
        for l in enriched:
            for k in greeks_net:
                v = l["greeks"].get(k)
                if v is not None:
                    greeks_net[k] += l["sign"] * l["quantity"] * float(v)

        def _payoff(s: float) -> float:
            total = 0.0
            for l in enriched:
                k = l["strike"]
                intrinsic = max(0.0, s - k) if l["type"] == "C" else max(0.0, k - s)
                total += l["sign"] * l["quantity"] * (intrinsic - l["mark_price"])
            return total

        lo = max(spot * 0.5, min(strikes) * 0.7) if strikes else spot * 0.5
        hi = max(spot * 1.5, max(strikes) * 1.3) if strikes else spot * 1.5
        steps = 14
        points = []
        for i in range(steps + 1):
            s = lo + (hi - lo) * (i / steps)
            points.append({"spot": round(s, 2), "pnl": round(_payoff(s), 4)})
        key_points = sorted({int(lo), int(spot), int(hi), *(int(k) for k in strikes)})
        for s in key_points:
            points.append({"spot": float(s), "pnl": round(_payoff(float(s)), 4)})
        points.sort(key=lambda p: p["spot"])

        pnls = [p["pnl"] for p in points]
        max_profit = max(pnls) if pnls else 0.0
        max_loss = min(pnls) if pnls else 0.0

        break_evens: list[float] = []
        for a, b in zip(points, points[1:], strict=False):
            if a["pnl"] == 0:
                break_evens.append(a["spot"])
            elif (a["pnl"] < 0 < b["pnl"]) or (a["pnl"] > 0 > b["pnl"]):
                frac = -a["pnl"] / (b["pnl"] - a["pnl"])
                break_evens.append(round(a["spot"] + frac * (b["spot"] - a["spot"]), 2))

        structure = self._guess_structure(enriched)

        sum(l["quantity"] * spot for l in enriched) if spot else 0.0
        fee_per_leg = min(0.0003 * (spot or 1) * sum(l["quantity"] for l in enriched),
                           0.125 * abs(net_premium)) if spot else 0.0
        fees_open = round(fee_per_leg, 4)
        fees_total = round(fees_open * 2, 4)

        warnings = []
        for l in enriched:
            if l["action"] == "short":
                any_long_same_type = any(
                    o["type"] == l["type"] and o["action"] == "long" and o["coin"] == l["coin"]
                    for o in enriched
                )
                if not any_long_same_type:
                    warnings.append(
                        f"naked short {l['type']} {l['name']} — viola hard prohibition"
                    )

        pl_ratio = (
            round(abs(max_profit / max_loss), 2)
            if max_loss and max_loss != 0 else None
        )

        return {
            "structure_name_guess": structure,
            "net_premium": round(abs(net_premium), 4),
            "net_premium_sign": "credit" if credit else "debit",
            "max_profit": round(max_profit, 4),
            "max_loss": round(max_loss, 4),
            "break_even": break_evens,
            "profit_loss_ratio": pl_ratio,
            "greeks_net": {k: round(v, 6) for k, v in greeks_net.items()},
            "fees_estimate": {
                "open": fees_open,
                "close_estimate": fees_open,
                "total": fees_total,
                "cap_hit": fees_total >= 0.125 * abs(net_premium) if net_premium else False,
            },
            "payoff_table": points,
            "spot_assumed": round(spot, 2),
            "warnings": warnings,
            "testnet": self.testnet,
        }

    @staticmethod
    def _guess_structure(legs: list[dict]) -> str:
        n = len(legs)
        if n == 1:
            l = legs[0]
            return f"long {l['type']}" if l["action"] == "long" else f"short {l['type']}"
        if n == 2:
            types = {l["type"] for l in legs}
            actions = {l["action"] for l in legs}
            strikes = sorted(set(l["strike"] for l in legs))
            if types == {"P"} and actions == {"long", "short"}:
                short = next(l for l in legs if l["action"] == "short")
                long_ = next(l for l in legs if l["action"] == "long")
                return "bull put spread" if short["strike"] > long_["strike"] else "bear put spread"
            if types == {"C"} and actions == {"long", "short"}:
                short = next(l for l in legs if l["action"] == "short")
                long_ = next(l for l in legs if l["action"] == "long")
                return "bear call spread" if short["strike"] < long_["strike"] else "bull call spread"
            if types == {"C", "P"} and len(strikes) == 1 and len(actions) == 1:
                return f"{list(actions)[0]} straddle"
            if types == {"C", "P"} and len(strikes) == 2 and len(actions) == 1:
                return f"{list(actions)[0]} strangle"
            if len({l["expiry"] for l in legs}) == 2 and len(strikes) == 1:
                return "calendar spread"
        if n == 4:
            types_list = [l["type"] for l in legs]
            if types_list.count("P") == 2 and types_list.count("C") == 2:
                return "iron condor"
        return "custom"

    async def find_by_delta(
        self,
        currency: str,
        expiry: str,
        target_delta: float,
        option_type: str,
        max_results: int = 3,
        min_open_interest: float = 100.0,
        min_volume_24h: float = 20.0,
    ) -> dict:
        import asyncio
        import datetime as _dt

        currency = currency.upper()
        option_type = option_type.lower()
        try:
            exp_dt = _dt.datetime.fromisoformat(expiry)
        except ValueError:
            exp_dt = _dt.datetime.strptime(expiry, "%Y-%m-%d")
        exp_ms = int(exp_dt.replace(tzinfo=_dt.UTC).timestamp() * 1000)
        day_ms = 24 * 3600 * 1000

        chain = await self.get_instruments(
            currency=currency,
            kind="option",
            expiry_from=expiry,
            expiry_to=(exp_dt + _dt.timedelta(days=1)).strftime("%Y-%m-%d"),
            limit=500,
            offset=0,
        )
        items = [
            i for i in chain.get("instruments", [])
            if i.get("option_type") == option_type
            and i.get("expiry") is not None
            and abs(i["expiry"] - exp_ms) < day_ms
        ]
        if not items:
            return {"matches": [], "best_match": None, "note": "no instruments for expiry"}

        async def _ticker(inst: str) -> dict:
            try:
                return await self.get_ticker(inst)
            except Exception:
                return {}

        tickers = await asyncio.gather(*[_ticker(i["name"]) for i in items])

        candidates = []
        for meta, tk in zip(items, tickers, strict=True):
            greeks = tk.get("greeks") or {}
            delta = greeks.get("delta")
            oi = meta.get("open_interest") or 0
            vol = tk.get("volume_24h") or 0
            bid = tk.get("bid") or 0
            ask = tk.get("ask") or 0
            if delta is None:
                continue
            if oi < min_open_interest or vol < min_volume_24h:
                continue
            mid = (bid + ask) / 2 if (bid and ask) else None
            spread_pct = round(100 * (ask - bid) / mid, 2) if mid and mid > 0 else None
            candidates.append({
                "instrument_name": meta["name"],
                "strike": meta["strike"],
                "delta": delta,
                "delta_distance": abs(delta - target_delta),
                "mark_iv": tk.get("mark_iv"),
                "mark_price_usd": tk.get("mark_price"),
                "bid_ask_spread_pct": spread_pct,
                "open_interest": oi,
                "volume_24h": vol,
                "greeks": greeks,
            })
        candidates.sort(key=lambda x: x["delta_distance"])
        top = candidates[:max_results]
        return {
            "currency": currency,
            "expiry": expiry,
            "target_delta": target_delta,
            "option_type": option_type,
            "matches": top,
            "best_match": top[0] if top else None,
            "candidates_considered": len(candidates),
            "testnet": self.testnet,
        }

    async def get_iv_rank(self, instrument: str) -> dict:
        currency = instrument.split("-", 1)[0].upper()
        if currency not in ("BTC", "ETH"):
            return {
                "instrument": instrument,
                "error": f"currency {currency} not supported for IV rank",
            }

        ticker_raw = await self._request(
            "public/ticker", {"instrument_name": instrument}
        )
        tr = ticker_raw.get("result") or {}
        current_iv = tr.get("mark_iv")

        import datetime as _dt

        now = _dt.datetime.now(_dt.UTC)
        series_by_lookback: dict[int, list[float]] = {}
        for lb in (30, 60, 90, 365):
            start = now - _dt.timedelta(days=lb)
            raw = await self._request(
                "public/get_volatility_index_data",
                {
                    "currency": currency,
                    "start_timestamp": int(start.timestamp() * 1000),
                    "end_timestamp": int(now.timestamp() * 1000),
                    "resolution": "1D",
                },
            )
            rows = (raw.get("result") or {}).get("data") or []
            series_by_lookback[lb] = [
                float(r[4]) for r in rows if len(r) >= 5
            ]

        if current_iv is None:
            latest = series_by_lookback.get(30) or series_by_lookback.get(90) or []
            current_iv = latest[-1] if latest else None

        def _percentile(values: list[float], target: float | None) -> float | None:
            if target is None or not values:
                return None
            below = sum(1 for v in values if v <= target)
            return round(100.0 * below / len(values), 2)

        def _rank(values: list[float], target: float | None) -> float | None:
            if target is None or not values:
                return None
            lo, hi = min(values), max(values)
            if hi == lo:
                return None
            return round(100.0 * (target - lo) / (hi - lo), 2)

        def _stats(values: list[float]) -> tuple[float | None, float | None]:
            if not values:
                return None, None
            m = sum(values) / len(values)
            var = sum((v - m) ** 2 for v in values) / len(values)
            return m, var ** 0.5

        mean_30, std_30 = _stats(series_by_lookback[30])
        warnings: list[str] = []
        if len(series_by_lookback[30]) < 10:
            warnings.append("limited_history")
        return {
            "instrument": instrument,
            "currency": currency,
            "current_iv": current_iv,
            "iv_percentile_30d": _percentile(series_by_lookback[30], current_iv),
            "iv_percentile_60d": _percentile(series_by_lookback[60], current_iv),
            "iv_percentile_90d": _percentile(series_by_lookback[90], current_iv),
            "iv_percentile_365d": _percentile(series_by_lookback[365], current_iv),
            "iv_rank_30d": _rank(series_by_lookback[30], current_iv),
            "iv_rank_60d": _rank(series_by_lookback[60], current_iv),
            "iv_rank_90d": _rank(series_by_lookback[90], current_iv),
            "iv_rank_365d": _rank(series_by_lookback[365], current_iv),
            "mean_30d": mean_30,
            "stddev_30d": std_30,
            "data_points_30d": len(series_by_lookback[30]),
            "data_timestamp": now.isoformat(),
            "warnings": warnings,
            "testnet": self.testnet,
        }

    async def get_realized_vol(
        self,
        currency: str,
        windows: list[int] | None = None,
    ) -> dict:
        """Annualized realized volatility (log-return std) su daily closes di Deribit index."""
        import datetime as _dt
        import math

        currency = currency.upper()
        if currency not in ("BTC", "ETH"):
            return {"currency": currency, "error": "currency not supported"}
        windows = windows or [14, 30]
        max_w = max(windows)

        now = _dt.datetime.now(_dt.UTC)
        start = now - _dt.timedelta(days=max_w + 10)
        raw = await self._request(
            "public/get_tradingview_chart_data",
            {
                "instrument_name": f"{currency}-PERPETUAL",
                "start_timestamp": int(start.timestamp() * 1000),
                "end_timestamp": int(now.timestamp() * 1000),
                "resolution": "1D",
            },
        )
        result = raw.get("result") or {}
        closes = [float(c) for c in (result.get("close") or [])]
        rv_by_window: dict[str, float | None] = {}
        for w in windows:
            if len(closes) <= w:
                rv_by_window[f"{w}d"] = None
                continue
            segment = closes[-(w + 1):]
            rets = [
                math.log(segment[i] / segment[i - 1])
                for i in range(1, len(segment))
                if segment[i - 1] > 0
            ]
            if len(rets) < 2:
                rv_by_window[f"{w}d"] = None
                continue
            m = sum(rets) / len(rets)
            var = sum((r - m) ** 2 for r in rets) / (len(rets) - 1)
            rv_by_window[f"{w}d"] = round(math.sqrt(var * 365) * 100, 2)

        iv_current = None
        try:
            dvol_raw = await self._request(
                "public/get_volatility_index_data",
                {
                    "currency": currency,
                    "start_timestamp": int((now - _dt.timedelta(days=2)).timestamp() * 1000),
                    "end_timestamp": int(now.timestamp() * 1000),
                    "resolution": "1D",
                },
            )
            dv = (dvol_raw.get("result") or {}).get("data") or []
            if dv:
                iv_current = float(dv[-1][4])
        except Exception:
            iv_current = None

        iv_rv_spread: dict[str, float | None] = {}
        for w_key, rv in rv_by_window.items():
            iv_rv_spread[w_key] = (
                round(iv_current - rv, 2) if (iv_current is not None and rv is not None) else None
            )

        return {
            "currency": currency,
            "realized_vol_pct": rv_by_window,
            "iv_current_pct": iv_current,
            "iv_minus_rv_pct": iv_rv_spread,
            "data_points": len(closes),
            "data_timestamp": now.isoformat(),
            "testnet": self.testnet,
        }

    async def get_dvol_history(
        self,
        currency: str,
        lookback_days: int = 90,
    ) -> dict:
        import datetime as _dt

        now = _dt.datetime.now(_dt.UTC)
        start = now - _dt.timedelta(days=lookback_days)
        raw = await self._request(
            "public/get_volatility_index_data",
            {
                "currency": currency.upper(),
                "start_timestamp": int(start.timestamp() * 1000),
                "end_timestamp": int(now.timestamp() * 1000),
                "resolution": "1D",
            },
        )
        rows = (raw.get("result") or {}).get("data") or []
        series = [
            {"timestamp": r[0], "dvol": float(r[4])}
            for r in rows
            if len(r) >= 5
        ]
        values = [s["dvol"] for s in series]
        values_sorted = sorted(values)

        def _pct(p: float) -> float | None:
            if not values_sorted:
                return None
            idx = int(round((len(values_sorted) - 1) * p))
            return values_sorted[idx]  # type: ignore[no-any-return]

        mean = sum(values) / len(values) if values else None
        return {
            "currency": currency.upper(),
            "lookback_days": lookback_days,
            "series": series,
            "current": values[-1] if values else None,
            "mean": mean,
            "p25": _pct(0.25),
            "p50": _pct(0.50),
            "p75": _pct(0.75),
            "p95": _pct(0.95),
            "data_points": len(values),
            "testnet": self.testnet,
        }

    async def get_technical_indicators(
        self,
        instrument: str,
        indicators: list[str],
        start_date: str,
        end_date: str,
        resolution: str = "1h",
    ) -> dict:
        historical = await self.get_historical(instrument, start_date, end_date, resolution)
        candles = historical.get("candles", [])
        closes = [c["close"] for c in candles]
        highs = [c["high"] for c in candles]
        lows = [c["low"] for c in candles]

        result: dict[str, Any] = {}
        for indicator in indicators:
            name = indicator.lower()
            if name == "sma":
                result["sma"] = ind.sma(closes, 20)
            elif name == "rsi":
                result["rsi"] = ind.rsi(closes)
            elif name == "atr":
                result["atr"] = ind.atr(highs, lows, closes)
            elif name == "macd":
                result["macd"] = ind.macd(closes)
            elif name == "adx":
                result["adx"] = ind.adx(highs, lows, closes)
            else:
                result[name] = None
        return result

    # ── Write tools ──────────────────────────────────────────────

    async def place_order(
        self,
        instrument_name: str,
        side: str,
        amount: float,
        type: str = "limit",
        price: float | None = None,
        reduce_only: bool = False,
        post_only: bool = False,
        label: str | None = None,
    ) -> dict:
        endpoint = "private/buy" if side == "buy" else "private/sell"
        params: dict[str, Any] = {
            "instrument_name": instrument_name,
            "amount": amount,
            "type": type,
        }
        if price is not None:
            if type in ("stop_market", "take_profit_market"):
                params["trigger_price"] = price
                params["trigger"] = "mark_price"
            else:
                params["price"] = price
        if label:
            params["label"] = label
        if reduce_only:
            params["reduce_only"] = True
        if post_only:
            params["post_only"] = True
        raw = await self._request(endpoint, params)
        r = raw.get("result")
        if r is None:
            return {"error": raw.get("error", "unknown"), "state": "error"}
        return r  # type: ignore[no-any-return]

    async def place_combo_order(
        self,
        legs: list[dict[str, Any]],
        side: str,
        amount: float,
        type: str = "limit",
        price: float | None = None,
        label: str | None = None,
    ) -> dict:
        """Crea un combo via private/create_combo poi piazza un singolo ordine
        (buy/sell) sull'instrument_name del combo. Una sola crociata di spread
        invece di N (uno per leg) → minor slippage su strutture liquide.

        legs: [{instrument_name, direction: 'buy'|'sell', ratio: int}].
        """
        combo_raw = await self._request("private/create_combo", {"trades": legs})
        combo = combo_raw.get("result")
        if combo is None:
            return {"state": "error", "error": combo_raw.get("error", "unknown")}
        combo_instrument = combo.get("instrument_name") or combo.get("id")
        order = await self.place_order(
            instrument_name=combo_instrument,
            side=side,
            amount=amount,
            type=type,
            price=price,
            label=label,
        )
        if order.get("state") == "error":
            return {"state": "error", "error": order.get("error"), "combo_instrument": combo_instrument}
        return {"combo_instrument": combo_instrument, **order}

    async def set_leverage(self, instrument_name: str, leverage: int) -> dict:
        """CER-016: pre-set account leverage per evitare default 50x testnet."""
        raw = await self._request(
            "private/set_leverage",
            {"instrument_name": instrument_name, "leverage": leverage},
        )
        if raw.get("result") is None:
            return {"state": "error", "error": raw.get("error", "unknown")}
        return {"state": "ok", "instrument": instrument_name, "leverage": leverage}

    async def cancel_order(self, order_id: str) -> dict:
        raw = await self._request("private/cancel", {"order_id": order_id})
        if raw.get("result") is None:
            return {
                "order_id": order_id,
                "state": "error",
                "error": raw.get("error", "unknown"),
            }
        r = raw["result"]
        return {
            "order_id": r.get("order_id"),
            "state": r.get("order_state"),
        }

    async def set_stop_loss(self, order_id: str, stop_price: float) -> dict:
        """
        Amend an existing order to add a stop-loss trigger via edit endpoint.
        Deribit does not have a standalone set_stop_loss; we use private/edit
        to update stop_price on the order.
        """
        raw = await self._request(
            "private/edit",
            {"order_id": order_id, "stop_price": stop_price},
        )
        if raw.get("result") is None:
            return {"order_id": order_id, "error": raw.get("error", "unknown")}
        r = raw["result"].get("order", raw["result"])
        return {
            "order_id": r.get("order_id"),
            "state": r.get("order_state"),
            "stop_price": r.get("stop_price"),
        }

    async def set_take_profit(self, order_id: str, tp_price: float) -> dict:
        """
        Amend an existing order to add a take-profit trigger via private/edit.
        """
        raw = await self._request(
            "private/edit",
            {"order_id": order_id, "stop_price": tp_price},
        )
        if raw.get("result") is None:
            return {"order_id": order_id, "error": raw.get("error", "unknown")}
        r = raw["result"].get("order", raw["result"])
        return {
            "order_id": r.get("order_id"),
            "state": r.get("order_state"),
            "tp_price": tp_price,
        }

    async def close_position(self, instrument_name: str) -> dict:
        raw = await self._request(
            "private/close_position",
            {"instrument_name": instrument_name, "type": "market"},
        )
        r = raw.get("result")
        if r is None:
            return {"instrument": instrument_name, "error": raw.get("error", "unknown"), "state": "error"}
        order = r.get("order", r)
        return {
            "order_id": order.get("order_id"),
            "state": order.get("order_state"),
        }