@pipeworx/kalshi

Connect: https://pipeworx.io/mcp — every tool in the catalog, including @pipeworx/kalshi’s. Install: one-click buttons

Connect to just the @pipeworx/kalshi pack

https://gateway.pipeworx.io/kalshi/mcp — only @pipeworx/kalshi’s own tools, nothing else in the catalog.

No MCP client? Skip the connection: POST https://gateway.pipeworx.io/v1/tools/search_packs {"query":"..."} to find a tool below, GET /v1/tools/<name> for its schema, POST the same URL with arguments for the data — see For AI agents.

Tools: 19

Market data from Kalshi, the CFTC-regulated US prediction-market exchange. Public reads need no credential; nothing in this pack ever touches an account.

Upstream: https://api.elections.kalshi.com/trade-api/v2 (docs: https://docs.kalshi.com/llms.txt, which serves every endpoint page as .md).

What is here

ToolWhat it answers
kalshi_markets / kalshi_marketfind markets by subject; one market in full, including its settlement rules
kalshi_events / kalshi_eventan event and every child market (one per outcome bucket)
kalshi_series / kalshi_macrothe recurring question templates; friendly-name shortcuts (Fed, CPI, BTC…)
kalshi_orderbook / kalshi_orderbooksdepth of book for one market / for up to 100 at once
kalshi_tradesexecuted trades, live or archived
kalshi_price_history / kalshi_candlesticksOHLC for one market with paging + archive fallback / for up to 100 on one shared window
kalshi_top_marketsmost-traded markets right now
kalshi_exchange_statusis the floor open
kalshi_event_live_datathe real-world measurement an event settles against
kalshi_game_statslive score + play-by-play for a sports milestone
kalshi_milestonesthe occurrences Kalshi hangs markets off, and which events belong to each
kalshi_structured_targetsthe entities (players, teams, actors, films, companies) markets refer to
kalshi_weather_index / kalshi_weather_index_calibrationsthe city temperature index temperature markets settle on, and how it is computed

Bold rows landed 2026-09-16 (fleet #2040). The bucket assignment for all eight probed endpoint families, with the literal responses, is in docs/pack-build-queue.md.

Traps already paid for

Kalshi filters Cloudflare egress on its listing endpoints. Several endpoints return empty to a Workers IP. The gateway injects _proxyUrl/_proxyToken and kalshiGet() routes every upstream fetch through the relay when they are present. Anything new must go through kalshiGet, never a bare fetch.

Auth-gated vs broken is legible on this API, so check before concluding. An endpoint that needs a credential answers 401 {"error":{"code":"token_authentication_failure"}}. A 400 bad_request is the handler refusing, not a missing key.

/markets/orderbooks takes tickers repeated; /markets/candlesticks takes market_tickers comma-separated. Two batch endpoints on the same API, two different array conventions. Getting it wrong is a 400, not a silent subset.

The batch candlestick response labels each group market_ticker, not ticker. Pair by that label. Pairing by request order would attribute one market’s entire price history to another and still return 200.

Kalshi caps batch candlesticks at 10,000 candles across ALL requested markets and gives no cursor there. kalshi_candlesticks reports truncated_by_upstream. For one ticker over a long window — or a market old enough to be archived — use kalshi_price_history, which pages and falls back to the archive; the batch endpoint does neither.

/live_data/events/{ticker} is 404 for most events, and that is normal. It exists for crypto price events and some economic-series events. found:false is the honest answer, not an error. A live BTC event’s payload is ~473 KB, which is why series are trimmed to max_points with series_trimmed reporting the untrimmed length.

Weather is city-keyed, not event-keyed. KXHIGHNY-26SEP16 404s on the event live-data endpoint. The observations live at /live_data/weather/{city} and exist whether or not a market is open.

There are exactly thirteen weather-index city ids, and they are ids, not place names: miami, dfw, houston, phl-delaware-valley, puget-sound, sf-bay, greater-boston, southeast-michigan, kansas-city, minneapolis-st-paul, nyc, chicago, la-coastal. Kalshi enumerates them in its own rejection message for an unknown city. austin is not one. The index is per-minute, so a window under ~2 minutes can legitimately return zero points. from/to here are unix milliseconds; last_sec is seconds.

Weather-index station offsets are in Celsius while the index itself is published in Fahrenheit. (kalshi_weather_index_calibrations.)

/live_data/milestone/{id}/game_stats answers 200 with {"pbp":{}} for a milestone Sportradar does not cover — verified on a college football game, 2026-09-16. plays_available:false says so rather than passing an empty array off as a game with no plays. Documented coverage: Pro/College Football, Pro/College Basketball, WNBA, Soccer, Pro Hockey, Pro Baseball.

/structured_targets does not reject an unknown type — it returns 200 with an empty list. The types are granular by sport; there is no bare player, team or athlete (type=player matches exactly one test row named testingmaxxiong). The handler names the known types whenever a filtered query comes back empty, so a typo cannot read as “Kalshi has no NFL players”.

/milestones requires limit (a 400 without it) and its unfiltered page skews heavily historical — pass minimum_start_date for current occurrences.

forecast_percentile_history validates and then refuses. Full evidence in docs/pack-build-queue.md; do not re-probe it looking for the right parameters.

Development

node scripts/sync-catalog.mjs          # after any tool change
pnpm check:examples                     # tool-examples.json must cover new tools

Smoke-test through the live gateway with the internal header, never with a BYO key (scripts/pwcall.sh call kalshi_weather_index '{"city":"nyc"}'), and check the payload is non-empty — a 200 with an empty array passes every CI gate and is still a failed smoke test.

Tools

  • kalshi_markets — Kalshi prediction markets matching a filter, each with its live Yes/No price.
  • kalshi_market — One Kalshi market in full, including the rules text that decides how it settles.
  • kalshi_events — Kalshi prediction market events matching a filter, each with its child markets, from Kalshi.
  • kalshi_event — A Kalshi event with every child market and its current odds, from the regulated US exchange.
  • kalshi_series — A Kalshi market series — a recurring question type — and its markets, from Kalshi.
  • kalshi_orderbook — The live bid and ask ladder for one Kalshi market, from Kalshi.
  • kalshi_trades — Trades recently executed on a Kalshi market, newest first, from Kalshi.
  • kalshi_price_history — How a Kalshi market’s Yes-price moved over time, as a timestamped series.
  • kalshi_exchange_status — Whether the Kalshi exchange is currently open for trading, from Kalshi.
  • kalshi_top_markets — The highest-volume open Kalshi markets right now, from Kalshi.
  • kalshi_macro — Kalshi’s macroeconomic indicator markets (CPI, Fed rate, jobs) and their live prices.
  • kalshi_orderbooks — Live bid/ask ladders for many Kalshi markets at once, from Kalshi’s batch endpoint.
  • kalshi_candlesticks — OHLC price history for many Kalshi markets on one shared timeline, from Kalshi.
  • kalshi_event_live_data — The observed underlying data behind a Kalshi event — price track, published series — from Kalshi.
  • kalshi_game_stats — Live score and play-by-play for a Kalshi sports milestone, from Kalshi (Sportradar-fed).
  • kalshi_milestones — Real-world occurrences Kalshi tracks — games, elections — and the events tied to each, from Kalshi.
  • kalshi_structured_targets — Kalshi’s entity index — players, teams, actors, films its markets reference — from Kalshi.
  • kalshi_weather_index — The multi-station city temperature index Kalshi settles temperature markets on, from Kalshi.
  • kalshi_weather_index_calibrations — Station weights, offsets and every recalibration behind a Kalshi city temperature index.

Tools

  • kalshi_candlesticks — OHLC price history for many Kalshi markets on one shared timeline, from Kalshi.
  • kalshi_event — A Kalshi event with every child market and its current odds, from the regulated US exchange.
  • kalshi_event_live_data — The observed underlying data behind a Kalshi event — price track, published series — from Kalshi.
  • kalshi_events — Kalshi prediction market events matching a filter, each with its child markets, from Kalshi.
  • kalshi_exchange_status — Whether the Kalshi exchange is currently open for trading, from Kalshi.
  • kalshi_game_stats — Live score and play-by-play for a Kalshi sports milestone, from Kalshi (Sportradar-fed).
  • kalshi_macro — Kalshi's macroeconomic indicator markets (CPI, Fed rate, jobs) and their live prices.
  • kalshi_market — One Kalshi market in full, including the rules text that decides how it settles.
  • kalshi_markets — Kalshi prediction markets matching a filter, each with its live Yes/No price.
  • kalshi_milestones — Real-world occurrences Kalshi tracks — games, elections — and the events tied to each, from Kalshi.
  • kalshi_orderbook — The live bid and ask ladder for one Kalshi market, from Kalshi.
  • kalshi_orderbooks — Live bid/ask ladders for many Kalshi markets at once, from Kalshi's batch endpoint.
  • kalshi_price_history — How a Kalshi market's Yes-price moved over time, as a timestamped series.
  • kalshi_series — A Kalshi market series — a recurring question type — and its markets, from Kalshi.
  • kalshi_structured_targets — Kalshi's entity index — players, teams, actors, films its markets reference — from Kalshi.
  • kalshi_top_markets — The highest-volume open Kalshi markets right now, from Kalshi.
  • kalshi_trades — Trades recently executed on a Kalshi market, newest first, from Kalshi.
  • kalshi_weather_index — The multi-station city temperature index Kalshi settles temperature markets on, from Kalshi.
  • kalshi_weather_index_calibrations — Station weights, offsets and every recalibration behind a Kalshi city temperature index.

Regenerated from source · build October 6, 2026