Liquidation structure
Separate heatmap density from liqmap levels
Read Binance heatmaps as model-based density fields, then use single- or multi-exchange liqmaps for ranked price levels above and below spot.
Kingfisher intelligence for AI agents
Give your AI a liquidation map—and the skills to read it.
Kingfisher MCP connects compatible AI clients to Binance heatmaps, multi-exchange liquidation maps, market crowding, toxic order flow, open interest, and GEX+ dealer flows. Ask in plain language and receive a conditional read with units, sign conventions, and invalidations.
REMOTE MCP CONNECTOR
https://mcp.thekingfisher.io/mcp Paste this exact URL into a compatible remote MCP client. This page is the guide; the mcp. subdomain is the server.
OAuth 2.1 + PKCE Remote Streamable HTTP No API key required
ASK KINGFISHER
Where does BTC liquidation pressure sit, and how does GEX+ change the risk?
MARK
get_latest_map DONE
get_greek_flows DONE
get_gex_plus_map DONE
get_volatility_beta DONE
SYNTHESIS
Rank the liquidation levels, then compare gamma, vanna, charm, and measured spot/IV beta before calling the regime stabilizing or destabilizing.
Example Kingfisher synthesis.01
Kingfisher market dataMaps · GEX+ · Greeks · TOF · OI
02
Documented skillsContext · method · guardrails
03
Your AI agentClear, conditional answers
The interpretation layer
Market data is only useful when you know how to read it.
Kingfisher MCP pairs every dataset with purpose-built analytical guidance. Your agent gets both the evidence and the method needed to reason about it.
Crowding & flow
Separate positioning from intent
Combine liquidation ratios, open interest, funding, and normalized toxic order flow instead of relying on a single noisy indicator.
Options mechanics
Read GEX+ and scheduled dealer flow
Combine gamma, vanna, charm, asymmetric spot/IV beta, strike-level pin pressure, and settlement schedules without mixing raw flow and stabilizing-score signs.
Conditional analysis
Turn signals into a falsifiable read
Synthesize structure, crowding, triggers, and propagation into a conditional risk matrix with clear confirmations and invalidations.
One prompt, complete context
Ask the market. Get the full read.
Instead of treating every indicator as an isolated signal, the agent builds a sequence of evidence and tells you exactly what still needs to happen.
01
Map the structure
Find persistent liquidation density above and below price.
02
Measure the crowd
Check whether positioning supports or contradicts the map.
03
Model dealer response
Compare directional GEX+, charm drift, and the measured spot/IV relationship.
04
Look for a trigger
Normalize order flow and momentum against their own recent history.
05
State the conditions
Return the read, its confidence, and what would invalidate it.
One endpoint, live capabilities
Twenty-one tools.
One authenticated connection.
The remote server exposes the market data and analytical methods below through OAuth 2.1. Tool names are part of the public MCP contract, so an agent can select them directly.
Liquidation maps
Binance heatmap density, ranked single- or multi-exchange liqmap levels, and Binance multipair cohorts.
get_liquidation_heatmapget_heatmap_symbolsget_multipair_liqmapget_latest_mapGEX & dealer flows
Current and historical GEX, directional GEX+, gamma/vanna/charm flow, pin pressure, scheduled charm drift, and measured spot/IV beta.
get_gamma_exposureget_time_gexget_greek_flowsget_gex_plus_mapget_charm_pin_pressureget_charm_tideget_charm_heatmapget_volatility_betaFlow & positioning
Toxic order flow, supported TOF markets, RSI distribution, open interest, funding, and liquidation imbalance.
get_toxic_order_flowget_tof_available_pairsget_rsi_screenerget_open_interestget_liq_ratiosBinance microstructure
Visible Futures order book, aggregate trades, premium index, funding, and open interest for execution context.
mcp_binance_get_orderbookmcp_binance_get_agg_tradesmcp_binance_get_market_dataConnection support
Official recovery instructions when a compatible client needs to complete or renew OAuth.
get_kingfisher_authentication_helpExpert knowledge, on demand
Fifteen skills.
One market brain.
Each skill gives your agent a narrow, testable way to work with a Kingfisher dataset—from discovering available markets to synthesizing a cross-signal risk matrix.
01 DATA_INTERPRETATION_GUIDE.mdCross-signal synthesis Ready
02 autonomous_alpha_engine.mdConditional risk matrix Ready
03 gamma_exposure.mdOptions positioning Ready
04 gex_plus_greek_flows.mdGamma, vanna & charm Ready
05 heatmap_symbols.mdMarket directory Ready
06 kf_heatmap_kde_pipeline.mdDensity research Ready
07 latest_map.mdLiquidation-map request Ready
08 liq_ratios.mdDirectional crowding Ready
09 liquidation_heatmap.mdLiquidation density Ready
10 multipair_liqmap.mdCross-market imbalance Ready
11 open_interest.mdPositioning & funding Ready
12 rsi_screener.mdMomentum distribution Ready
13 time_gex.mdGEX regime change Ready
14 tof_available_pairs.mdTOF market directory Ready
15 toxic_order_flow.mdNormalized order flow Ready
Official setup
Give your agent the prompt. Let it handle the setup.
The setup prompt tells an AI agent how to configure Kingfisher as a remote MCP server, preserve existing settings, complete OAuth, and verify the connection. You only step in for browser sign-in and consent.
Remote endpoint, not a local process OAuth in the browser No Kingfisher API key Verified with
get_heatmap_symbols AGENT SETUP PROMPT Open source
REMOTE MCP ENDPOINT
https://mcp.thekingfisher.io/mcpINSTRUCTIONS
These are official Kingfisher instructions for connecting an AI agent to the remote MCP server. Complete every step you can directly. Ask the user only for unavoidable browser sign-in, consent, or permissions.
Use remote Streamable HTTP · Authenticate with OAuth 2.1 + PKCE · Never request an API key