Risk & P&L - Mithril

Ask a prediction-market trader their exposure to one event and most reach for a spreadsheet. Mithril answers directly — per subaccount or across the whole workspace, with venues netted into one view.

Risk statistics

# Per subaccount
curl https://api.trymithril.com/v1/risk/stats/sub_... \
  -H "Authorization: Bearer $MITHRIL_API_KEY"

# Workspace-wide
curl https://api.trymithril.com/v1/risk/stats \
  -H "Authorization: Bearer $MITHRIL_API_KEY"

stats = s.get(f"{BASE}/v1/risk/stats/{sub_id}").json()
print(stats["gross_exposure"], stats["unrealized_pnl"], stats["largest_position_pct_bps"])

{
  "workspace_id": "ws_...",
  "subaccount_id": "sub_...",
  "gross_exposure": "1840.00",
  "market_value": "1902.50",
  "marks_stale": false,
  "unrealized_pnl": "62.50",
  "realized_pnl_today": "18.20",
  "realized_pnl_total": "412.90",
  "open_order_count": 3,
  "resting_notional": "250.00",
  "position_count": 7,
  "market_count": 5,
  "largest_position_pct_bps": 3200,
  "per_market_exposure": [ { "market_id": "mkt_...", "exposure": "588.00", "pct_bps": 3200 } ],
  "limit_utilization": {
    "max_total_notional": { "limit": "5000", "used": "1840.00", "pct_bps": 3680 }
  },
  "kill_switch": false,
  "as_of": "2026-07-16T14:32:00Z"
}

Percentages are in basis points (largest_position_pct_bps: 3200 = 32%). marks_stale: true warns that mark prices couldn’t be refreshed, so mark-to-market values are best-effort.

Daily P&L history

curl https://api.trymithril.com/v1/risk/pnl_history \
  -H "Authorization: Bearer $MITHRIL_API_KEY"

{
  "workspace_id": "ws_...",
  "subaccount_id": "sub_...",
  "days": 30,
  "points": [
    { "date": "2026-07-16", "realized_pnl": "18.20", "fees": "1.05",
      "volume": "540.00", "fill_count": 6, "cumulative_realized_pnl": "412.90" }
  ],
  "total_realized_pnl": "412.90",
  "total_fees": "27.40",
  "total_volume": "18240.00",
  "total_fills": 214,
  "as_of": "2026-07-16T14:32:00Z"
}

Fills

Every execution, newest first — with fees computed from each venue’s schedule:

curl "https://api.trymithril.com/v1/fills?limit=50" \
  -H "Authorization: Bearer $MITHRIL_API_KEY"

fills = s.get(f"{BASE}/v1/fills", params={"limit": 50}).json()["data"]
for f in fills:
    print(f["side"], f["quantity"], "@", f["price"], "fee", f["fee"])

Each fill (FillView):

{
  "id": "fill_...",
  "subaccount_id": "sub_...",
  "order_id": "ord_...",
  "exchange": "polymarket",
  "market_id": "mkt_...",
  "token_id": "7194...",
  "side": "yes",
  "action": "buy",
  "price": "0.43",
  "quantity": "125",
  "fee": "1.07",
  "realized_pnl": "0",
  "created_at": "2026-07-16T14:30:05Z"
}

Positions and portfolio

GET /v1/positions returns a paginated list of PositionView:

{
  "data": [
    {
      "id": "pos_...",
      "subaccount_id": "sub_...",
      "exchange": "polymarket",
      "market_id": "mkt_...",
      "token_id": "7194...",
      "side": "yes",
      "quantity": "125",
      "cost_basis": "53.75",
      "avg_price": "0.43",
      "mark_price": "0.44",
      "marked_at": "2026-07-16T14:32:00Z",
      "created_at": "2026-07-16T14:30:05Z",
      "updated_at": "2026-07-16T14:32:00Z"
    }
  ],
  "has_more": false
}

GET /v1/portfolio aggregates positions for a scope (PortfolioView):

{
  "workspace_id": "ws_...",
  "subaccount_id": "sub_...",
  "position_count": 7,
  "total_cost_basis": "1840.00",
  "positions": [ /* ...PositionView... */ ]
}

Mark-to-market value and P&L live on risk stats; the portfolio view is positions + aggregate cost basis.

Two more reporting endpoints:

Endpoint	Returns
`GET /v1/receipts`	Execution receipts (arrival vs achieved vs naive), one per worked order
`GET /v1/analytics/execution`	Predicted-vs-realized analytics per fill: `arrival_mid`, `predicted_avg`, `realized_avg`, `slippage_bps`, `prediction_error_bps`

curl https://api.trymithril.com/v1/portfolio \
  -H "Authorization: Bearer $MITHRIL_API_KEY"

python 06_risk.py walks all of these in one run — a good template for a monitoring dashboard or an end-of-day report.

Full endpoint reference

Every path, method, and auth requirement.

​Risk statistics

​Daily P&L history

​Fills

​Positions and portfolio

Full endpoint reference

Risk statistics

Daily P&L history

Fills

Positions and portfolio