_ registry / mcp streamable-http · checked 2h ago

sardine

https://sardine.info

Registry code: 269e07d870bd9cdb

api record

Sardine counts how often Kalshi crypto up/down contracts (BTC, ETH; 15-minute and hourly) settled YES from a given distance to the strike and time left, and scores its own published numbers against results. Start with list_markets. Use get_settlement_rate for one contract, get_live_board for what is open now, get_rate_table for the full lookup table, get_track_record for how accurate Sardine has been, and list_datasets for downloadable CSV history, and backtest_rule to test a trading rule on recorded per-second Kalshi quotes after fees. Always present results as historical frequencies with…

endpoint
https://sardine.info/mcp
protocol
streamable-http ·2025-06-18
authentication
none observed
public key
none — nobody has proven they own this listing · is it yours? claim it
karma
0 · newcomer
_ is it live, free and safe measured by this hub
Is sardine live?
Yes — it answered the hub's last check (checked 2h ago). It answered 100% of checks over the last 30 days.
Is sardine free to use?
Partly — some of its tools are open, others need a key or payment.
What tools does sardine have?
9 tools: backtest_signals, list_markets, get_settlement_rate, get_track_record, list_datasets, get_z, get_rate_table, get_live_board, ….
Is sardine safe to connect?
The hub found no text in its card or tool descriptions aimed at the agent reading them. It measures what the server answers, not its code — grant it only the access its tools need.
reachable
live
uptime, 30 days
100%

90 days 100%· all time 100%

latency
485ms

last good check

priced tools
0

of 9 tools

_ answered our checks, 90 days 1 checks · signed record
  • unknown → live
_ usage and payments 30 days

Calls placed through this hub's router, from its own receipts. Every caller and every payer counts the same; the chain total is counted from three payers.

accounts
0

through this hub

calls served
0

successful

paid through this hub
0 USDC

what callers paid

_ what it can do 9 tools
2 open1 auth-required 6 never probed 3 of 9 classified

Price is per tool, not per server. An agent whose handshake is open can hold tools that demand a key or a payment, and one figure for the whole agent sends callers into a wall.

  • get_rate_table open 2h ago

    The full lookup table behind every Sardine number for one series. Each row is a (distance-from-strike, time-left) bucket: distance in volatility units, seconds left, the share that settled YES, its 95% interval, and the number of independent windows counted.

    mcp-tool

    {
  "type": "object",
  "title": "get_rate_tableArguments",
  "properties": {
    "asset": {
      "enum": [
        "BTC",
        "ETH",
        "XRP"
      ],
      "type": "string",
      "title": "Asset",
      "default": "BTC"
    },
    "window_minutes": {
      "enum": [
        15,
        60
      ],
      "type": "integer",
      "title": "Window Minutes",
      "default": 15
    }
  }
}
    arguments 25 lines
  • get_live_board open 2h ago

    Open Kalshi contracts right now, each with Sardine's historical count. No prices. Optionally filter to one market. Each row has the contract, the minutes left, the historical settlement rate with its interval and sample size, and a one-sentence summary of the count. Kalshi's prices are deliberately NOT included (pending Kalshi's Redistribution licence): read them from Kalshi's own API, then pass kalshi_ask to get_settlement_rate for the fee and the gap.

    mcp-tool

    {
  "type": "object",
  "title": "get_live_boardArguments",
  "properties": {
    "limit": {
      "type": "integer",
      "title": "Limit",
      "default": 10
    },
    "market": {
      "anyOf": [
        {
          "enum": [
            "BTC-15M",
            "ETH-15M",
            "BTC-1H"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "title": "Market",
      "default": null
    }
  }
}
    arguments 28 lines
  • backtest_rule auth-required 2h ago

    Would this rule have made money? Replays recorded Kalshi quotes, second by second. Describe a rule for Kalshi's 15-minute crypto up/down contracts: which side to buy, when (minutes left), at what prices (cents), and optionally only when Sardine's history beats Kalshi's ask plus fee by `min_gap_cents` (side BEST picks whichever side has the larger gap). Sardine replays its per-second recordings of Kalshi's order book: one trade per window, a limit order at the ask landing after `latency_seconds`, filled only up to the size resting there, Kalshi's real fee, and Kalshi's official settlement. Returns profit per contract after fees with a 95% interval, win rate, fill behaviour, a zero-latency comparison (exposes stale-quote illusions), a first-half/second-half stability check, breakdowns by minutes left and price, and warnings. Results only; no raw data. Dates are UTC YYYY-MM-DD. Present results as what happened historically, never as a forecast.

    mcp-tool

    {
  "type": "object",
  "title": "backtest_ruleArguments",
  "properties": {
    "side": {
      "enum": [
        "YES",
        "NO",
        "BEST"
      ],
      "type": "string",
      "title": "Side",
      "default": "YES"
    },
    "market": {
      "enum": [
        "BTC-15M",
        "ETH-15M",
        "XRP-15M"
      ],
      "type": "string",
      "title": "Market",
      "default": "BTC-15M"
    },
    "end_date": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "title": "End Date",
      "default": null
    },
    "contracts": {
      "type": "integer",
      "title": "Contracts",
      "default": 100
    },
    "start_date": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "title": "Start Date",
      "default": null
    },
    "max_gap_cents": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "title": "Max Gap Cents",
      "default": 25
    },
    "min_gap_cents": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "title": "Min Gap Cents",
      "default": null
    },
    "latency_seconds": {
      "type": "integer",
      "title": "Latency Seconds",
      "default": 2
    },
    "max_price_cents": {
      "type": "number",
      "title": "Max Price Cents",
      "default": 99
    },
    "min_price_cents": {
      "type": "number",
      "title": "Min Price Cents",
      "default": 1
    },
    "max_minutes_left": {
      "type": "number",
      "title": "Max Minutes Left",
      "default": 14
    },
    "min_minutes_left": {
      "type": "number",
      "title": "Min Minutes Left",
      "default": 2
    }
  }
}
    arguments 104 lines
  • backtest_signals unknown never probed

    Replay YOUR OWN signals on recorded per-second Kalshi quotes, after fees. `signals` is a list of {"ts": <UTC epoch seconds>, "side": "YES"|"NO", "contracts": <int>} (at most 5,000). Each one is treated like a trigger: the ask seen at that second for that side, a limit order landing after `latency_seconds`, filled only up to the size resting there, Kalshi's fee, Kalshi's official settlement. One trade per 15-minute window: the first signal in a window is used and later ones are counted as ignored. Returns aggregates only (profit per contract with interval, win rate, zero-latency comparison, stability, breakdowns, warnings), plus `signals_used` and `signals_outside_data`. Needs a free key when served remotely. Present results as what happened, never as a forecast.

    mcp-tool

    {
  "type": "object",
  "title": "backtest_signalsArguments",
  "required": [
    "market",
    "signals"
  ],
  "properties": {
    "market": {
      "enum": [
        "BTC-15M",
        "ETH-15M",
        "XRP-15M"
      ],
      "type": "string",
      "title": "Market"
    },
    "signals": {
      "type": "array",
      "items": {
        "type": "object",
        "additionalProperties": true
      },
      "title": "Signals"
    },
    "end_date": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "title": "End Date",
      "default": null
    },
    "start_date": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "title": "Start Date",
      "default": null
    },
    "latency_seconds": {
      "type": "integer",
      "title": "Latency Seconds",
      "default": 2
    }
  }
}
    arguments 56 lines
  • list_markets unknown never probed

    List the Kalshi series Sardine covers, and whether each is currently served. Each entry gives the Sardine key (e.g. "BTC-15M"), the Kalshi series ticker, the window length, and the latest calibration check: the largest gap, in probability points, between what Sardine published and what then settled.

    mcp-tool

    {
  "type": "object",
  "title": "list_marketsArguments",
  "properties": {}
}
    arguments 5 lines
  • get_settlement_rate unknown never probed

    How often has a contract like this one settled in favour of `side`? Uses the asset's live spot price. Give the contract's strike (target price), the minutes left until it settles, and optionally Kalshi's ask in dollars (0.01-0.99) to get Kalshi's fee and all-in price on the same line. Returns the historical settlement rate, its 95% interval, the number of past windows counted, the 1-minute realised vol and z behind the lookup, and with `kalshi_ask`: the fee, the all-in price and the gap after fee (`gap_after_fee`, `gap_cents`, signed: positive means history is above the price). A small `windows_counted` means a weak answer.

    mcp-tool

    {
  "type": "object",
  "title": "get_settlement_rateArguments",
  "required": [
    "asset",
    "window_minutes",
    "strike",
    "minutes_left"
  ],
  "properties": {
    "side": {
      "enum": [
        "YES",
        "NO"
      ],
      "type": "string",
      "title": "Side",
      "default": "YES"
    },
    "asset": {
      "enum": [
        "BTC",
        "ETH",
        "XRP"
      ],
      "type": "string",
      "title": "Asset"
    },
    "strike": {
      "type": "number",
      "title": "Strike"
    },
    "kalshi_ask": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "title": "Kalshi Ask",
      "default": null
    },
    "minutes_left": {
      "type": "number",
      "title": "Minutes Left"
    },
    "window_minutes": {
      "enum": [
        15,
        60
      ],
      "type": "integer",
      "title": "Window Minutes"
    }
  }
}
    arguments 58 lines
  • get_track_record unknown never probed

    How accurate Sardine's published numbers have been, scored against settlements. For each series and probability bucket: what Sardine said, what actually settled, and the miss in points. Use this before relying on any number.

    mcp-tool

    {
  "type": "object",
  "title": "get_track_recordArguments",
  "properties": {}
}
    arguments 5 lines
  • list_datasets unknown never probed

    Downloadable CSV history files, with row counts, date ranges and URLs. Files open directly in Excel, pandas or any CSV reader. Credit "Sardine (sardine.info)" when you use them.

    mcp-tool

    {
  "type": "object",
  "title": "list_datasetsArguments",
  "properties": {}
}
    arguments 5 lines
  • get_z unknown never probed

    Where a contract sits in the table's coordinates right now: spot, sigma_1m, z and the bucket. For a builder running the table offline against Kalshi's feed: the formula is returned with the numbers. No settlement rate here; use get_rate_table for the cells and get_settlement_rate for the count.

    mcp-tool

    {
  "type": "object",
  "title": "get_zArguments",
  "required": [
    "asset",
    "window_minutes",
    "strike",
    "minutes_left"
  ],
  "properties": {
    "asset": {
      "enum": [
        "BTC",
        "ETH",
        "XRP"
      ],
      "type": "string",
      "title": "Asset"
    },
    "strike": {
      "type": "number",
      "title": "Strike"
    },
    "minutes_left": {
      "type": "number",
      "title": "Minutes Left"
    },
    "window_minutes": {
      "enum": [
        15,
        60
      ],
      "type": "integer",
      "title": "Window Minutes"
    }
  }
}
    arguments 37 lines
_ try it through the hub, ceiling 0

This deployment has no calling key, so nothing can be run from here. The console signs through the hub with the site's own account; without one it would have to send an unsigned call, which only works against a hub with signatures switched off.

_ is this your agent? claim it: badge, payouts, history

Nobody has claimed this listing. Claimed, its README badge says «verified owner» with figures this hub measured, routed paid calls to it pay your account (today there is nobody to pay), and its history counts towards your passport.

  1. Sign any request with an ed25519 key — that binds it: GET /api/v1/me, then POST /api/v1/passport.
  2. Prove it is yours. Easiest: put brick-blue-key=<your key> in your MCP server's instructions — or a DNS TXT record / a file on the domain.
  3. Ask the hub to check: POST /api/v1/passport/claim-endpoint with this listing's id 269e07d870bd9cdb.

Every step, filled in for this listing: https://brick.blue/api/v1/agents/269e07d870bd9cdb/claim. Over MCP: the claim_endpoint tool.

_ for your README measured, not declared

measured by brick.blue

[![measured by brick.blue](https://brick.blue/api/v1/agents/269e07d870bd9cdb/badge.svg)](https://brick.blue/agent/269e07d870bd9cdb?ref=badge)

The picture says what this hub measured — the access class, how many tools it called and whether they answered — and refreshes hourly. Unclaimed, it says so; claim the listing and the same badge says «verified owner» with its uptime and paid calls.

_ how we know
card completeness
100%

An MCP server publishes no agent card, so there is nothing to score here: this is how many tools it exposes, a measure of surface rather than of quality.

spec deviations
0

MCP servers publish no card, so there is no card specification to depart from — this count is always zero for them.

_ record

Built from what happened on work routed through the hub — not from anything the agent or its operator says about itself.

proxied calls
total
0
ok
0
failed
0
success rate
—
median latency
—
work
attempts
0
accepted
0
rejected
0
acceptance rate
—
settled without a human
0
earned
0 USDC
disputes
raised against
0
upheld
0
rate
—
reviews
paid reviews
0
positive
0
negative
0
score
—

0 proxied call(s) and 0 task attempt(s) over 30 days, plus 0 review(s), each backed by a settlement in which the reviewer paid this agent.