advanced_web_research
ChatGPTUse this when the user asks to research recent company announcements or public-web news and summarize relevant sources. Returns web evidence with source URLs, titles and available publication dates. Optional domain and date filters narrow the search. Research MCP defaults to 5 results and 2000 text characters per result; oversized evidence excerpts may be shortened with an explicit notice. Use focused searches for additional detail. exa connector tool exa_web_search_advanced_exa; capabilities: market.news, market.macro, crypto.news. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
ask_jack
ChatGPTAsk StableJack Navigator a research question and receive its synthesized answer using the normal exchange-chat experience. This creates or reuses a persistent Navigator conversation and uses normal Navigator credit billing.
asset_brief_research_packet
ChatGPTSelected-skill-only research packet for asset-brief. Normalize the user request into one typed intent, then call exactly once. The intent accepts one asset or up to three independent assets, each with a model-classified asset class and any explicit instrument preference. The packet owns identity and instrument resolution, date windows, provider fan-out, retry, closed-candle calculations, typed gaps, and resumable clarification repair. Form the final prose brief only from its This call consumes 25 StableJack credits before execution, including when provider execution later fails.
batch_financial_statements
ChatGPTFetch one to twelve authorized public-company financial statements in a single prepared call. Use requests for independent income statement, balance sheet, or cash-flow reads, with one shared annual or quarterly period and bounded row limit. Put all independent statement reads needed for the answer into one call; results preserve request order and report a typed outcome per item. This tool can only invoke the exact request-authorized FMP statements descriptor and cannot selec This call consumes 25 StableJack credits before execution, including when provider execution later fails.
batch_market_snapshot
ChatGPTFetches a compact live snapshot for 1-8 Hyperliquid markets in one tool call. Use this when one or more markets need multiple current fields across price action, funding, volume, open interest, spread, or visible order-book liquidity. Prefer it over repeated single-purpose reads when its compact fields cover the request. Market and L2 requests run concurrently, and the response summarizes depth without returning verbose raw levels. Inspect each markets row: status=error is no This call consumes 25 StableJack credits before execution, including when provider execution later fails.
batch_research
ChatGPTExecute two to eight independent request-authorized, system-access, read-only tools in one call after their exact schemas are visible. Copy each normal tool name and input unchanged into calls; all calls start together and results preserve call order. Use this instead of making sequential calls when no call depends on another call result. User-scoped MCP tools, including web and research tools, must be called directly rather than placed in this batch. Do not use it for discov This call consumes 25 StableJack credits before execution, including when provider execution later fails.
blockchain_intelligence
ChatGPTReturns a combined chain-level analytics snapshot from DefiLlama in a SINGLE call. Always fetches chain info (TVL, 7-day trend, bridge flows, stablecoins, assets) AND metrics (fees, DEX volume) together. Optionally include "options" in metricKinds if relevant. HISTORICAL TVL (set includeTvlHistory=true): attaches info.tvlHistory with the FULL daily chain TVL history. Supports ISO fromDate/toDate windows, lastN for recent slice (up to ~4000 days back), and `maxSe This call consumes 25 StableJack credits before execution, including when provider execution later fails.
buffett_research_packet
ChatGPTSelected-skill-only packet for analyze-like-buffett. Call exactly once with one classified tradable asset and horizon. It owns identity resolution and bounded market, annual financial, ownership, transcript and catalyst reads. Exact URL lookup is Codex native web only and is passed as a receipt. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
burry_research_packet
ChatGPTSelected-skill-only packet for analyze-like-burry. Call exactly once with one classified tradable asset and horizon. It owns identity resolution and bounded market, annual financial, dilution, ownership, transcript and catalyst reads. Exact URL lookup is Codex native web only and is passed as a receipt. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
centralized_exchange_volumes
ChatGPTDefiLlama public centralized-exchange (CEX) venue metrics only: spot volume, reported reserve assets, clean assets, and 24h/7d/30d venue inflows. Use for comparisons or questions about named venues such as Binance or Coinbase. Do not use for DEX/perpetual volume, open interest, live asset prices, corporate financials, order execution, or balances and holdings in the user's authenticated exchange account. --- Native skill: defi-data --- --- name: defi-data description: Core This call consumes 25 StableJack credits before execution, including when provider execution later fails.
crowd_positioning
ChatGPTAnalyzes crowd positioning on Hyperliquid. Two modes: (1) Market-wide: overall long/short counts, open interest trends, and per-coin heatmap. (2) Per-coin: positioning metrics for a specific coin, optionally scoped to a crowd segment. For an exact finite universe, pass all qualified symbols through coins; the result reports requested, matched, and missing coverage and does not apply the default top-exposure cut. Use this for questions like "how is the crowd positioned?", "is This call consumes 25 StableJack credits before execution, including when provider execution later fails.
crypto_etf_flows
ChatGPTDefiLlama Bitcoin or Ethereum spot-ETF flows only: asset-level dated net-flow history plus current fund AUM, volume, fee, and flow snapshots. This is the primary ETF-flow source. A ticker filter narrows the current fund snapshot; historical rows remain explicitly asset-level. Do not use for equity ETF holdings, prices, news, or trade execution. --- Native skill: defi-data --- --- name: defi-data description: Core reference for DefiLlama MCP tools. Maps DeFi questions to the This call consumes 25 StableJack credits before execution, including when provider execution later fails.
crypto_insight_by_symbol
ChatGPTGet the current StableJack crypto AI insight record for one symbol, optionally scoped to SPOT or PERPS. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
crypto_insights
ChatGPTGet all current StableJack crypto AI insight records, optionally filtered by SPOT or PERPS venue. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
crypto_market_data
ChatGPTFetches current crypto prices, market caps, and 24h changes from CoinGecko. Use topN for a current market-cap ranking, or assetIds for specific assets. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
defi_category_metrics
ChatGPTDefiLlama DeFi category aggregates and rankings only, such as Lending or DEX category TVL, fees, protocol count, and category TVL history. Do not use for a named protocol or chain, token prices, CEXs, or public-company sectors and financials. --- Native skill: defi-data --- --- name: defi-data description: Core reference for DefiLlama MCP tools. Maps DeFi questions to the correct tool call with proper parameters. Covers entity conventions, metric interpretation, stock vs fl This call consumes 25 StableJack credits before execution, including when provider execution later fails.
defi_market_tvl
ChatGPTDefiLlama global aggregate DeFi TVL and its dated history only. Use for total market-wide DeFi TVL, change, or trend. Do not use for a named protocol, chain, token price, trading volume, category ranking, or public-company data; use the matching narrow tool instead. --- Native skill: defi-data --- --- name: defi-data description: Core reference for DefiLlama MCP tools. Maps DeFi questions to the correct tool call with proper parameters. Covers entity conventions, metric int This call consumes 25 StableJack credits before execution, including when provider execution later fails.
defi_yield_pools
ChatGPTDefiLlama DeFi yield-pool and lending/borrowing rate screening only: pool APY, base/reward APY, borrow rate, TVL, chain, protocol, and token filters. Do not use for token price, generic protocol metrics, CEX earn products, or investment-performance forecasts. --- Native skill: defi-data --- --- name: defi-data description: Core reference for DefiLlama MCP tools. Maps DeFi questions to the correct tool call with proper parameters. Covers entity conventions, metric interpreta This call consumes 25 StableJack credits before execution, including when provider execution later fails.
druckenmiller_research_packet
ChatGPTSelected-skill-only packet for analyze-like-druckenmiller. Call exactly once with one classified tradable asset and horizon. It owns identity resolution, bounded class-specific market, financial, macro, positioning and catalyst reads, batching, retries, typed gaps, timestamps, and provenance. The model applies the Druckenmiller framework itself. It never calls Exa, Grok/xAI, Perplexity, or another model-backed search path; exact URL lookup is Codex native web only and is pass This call consumes 25 StableJack credits before execution, including when provider execution later fails.
equity_symbol_discovery
ChatGPTUse this to resolve an ambiguous company name or discover a canonical equity symbol and venue. For a current quote on an already known ticker, use market_quotes. Discovery metadata is not a substitute for a timestamped market quote. Required when a public-company read still needs identity resolution: when the user supplies one or more unresolved company names, stock/equity symbols, or ticker-like identifiers that are not known crypto, call this tool in the same turn before reading by ticker or giving the final answer. Do not repeat identity discovery for symbols already resolved by an authenticated structured provider result in the same turn. It resolves canonical FMP tickers and can return batched curre This call consumes 25 StableJack credits before execution, including when provider execution later fails.
funding_rate_history
ChatGPTRetrieves historical funding rate records for a single Hyperliquid perp symbol. Returns timestamped funding rate + premium entries between startTime and optional endTime. Times are in milliseconds since epoch. Window boundaries are start-inclusive, end-exclusive: [startTime, endTime). Supports HIP-3 builder DEX assets via dex param or namespace syntax (e.g. xyz:TSLA). This call consumes 25 StableJack credits before execution, including when provider execution later fails.
historical_market_prices
ChatGPTUnified price lookup tool — handles live, instant, date-based, range, and extremes price queries. CRITICAL: For relative time ("3 days ago", "yesterday", "2 hours ago"), pass daysAgo or hoursAgo as a NUMBER. The tool resolves the actual date server-side using the real server clock. Do NOT compute ISO dates yourself. Modes: • mode: "live" — current/live price (markPrice). No time fields needed. • mode: "instant" — price at a specific moment. Use hoursAgo (e.g. 2) or targetTi This call consumes 25 StableJack credits before execution, including when provider execution later fails.
l2_order_book
ChatGPTFetches the L2 order book snapshot for a single Hyperliquid perp symbol. Returns bids and asks with price, size, and order count per level. Also computes best bid/ask, spread, and spread in basis points. Optional nSigFigs (2-5) aggregates price levels; mantissa (2 or 5) further controls buckets when nSigFigs=5. Optional maxLevels (default 20) limits depth per side to keep response compact. Supports HIP-3 builder DEX assets via dex param or namespace syntax (e.g. xyz:TSLA). This call consumes 25 StableJack credits before execution, including when provider execution later fails.
liquidation_clusters_research_packet
ChatGPTSelected-skill-only packet for liquidation-clusters. Call exactly once with an asset or scan intent. For a named asset, set mode="asset", supply market, and omit shortlistSize. For a broad scan, set mode="scan" and use shortlistSize 1-8 (normally 6). If rangePct is supplied it must be 0.01-0.5 (normally 0.15). It owns Hyperliquid identity resolution, bounded market screening, position-level liquidation reads, aggregate risk corroboration, optional thin-book depth, determinist This call consumes 25 StableJack credits before execution, including when provider execution later fails.
liquidation_risk
ChatGPTAnalyzes liquidation risk on Hyperliquid. Two modes: (1) Market-wide risk overview: per-coin liquidation risk table showing which coins have the most value at risk. (2) Coin-specific heatmap: zone-aggregated liquidation clusters for a given coin, with top clusters and above/below totals. Use this for "where are liquidations clustered?", "what coins have the highest liq risk?", "show me BTC liquidation levels". --- Native skill: SKILL.md --- HyperTracker API Pre-computed This call consumes 25 StableJack credits before execution, including when provider execution later fails.
market_calculator
ChatGPTServer-owned pair-statistics calculator for normal chat and the fundamental-pair-trading skill. Use it for exact correlation, beta, per-leg volatility, beta-hedged spread volatility, ratio statistics, a 60-observation rolling spread z-score, and an Engle-Granger residual ADF cointegration result between two assets. Call it once with required assetClass, the two ticker symbols, and a structured timeWindow. assetClass must match both legs: equity for two listed-company/equity-p This call consumes 25 StableJack credits before execution, including when provider execution later fails.
market_candle_data
ChatGPTFetches OHLCV candle data for a single Hyperliquid perp symbol. Returns open, high, low, close, volume, and trade count per interval. Supported intervals: 1m, 3m, 5m, 15m, 30m, 1h, 2h, 4h, 8h, 12h, 1d, 3d, 1w, 1M. Times include milliseconds since epoch plus server-derived openTimeIso/closeTimeIso strings; use those exact ISO timestamps for labels and never synthesize later historical dates. For relative phrases, do not compute dates yourself: use periodHours, periodDays, or s This call consumes 25 StableJack credits before execution, including when provider execution later fails.
market_segment_bias
ChatGPTAnalyzes how a specific crowd segment (whales, smart money, shrimp, etc.) is positioned over time on Hyperliquid. Returns bias direction, exposure trend, and time-series history. Accepts segment names directly — no need to call a separate segment-discovery tool first. --- Native skill: SKILL.md --- HyperTracker API Pre-computed intelligence layer for Hyperliquid, the largest on-chain perpetual futures exchange. Cohort analytics, live order flow, liquidation risk scoring, This call consumes 25 StableJack credits before execution, including when provider execution later fails.
news_brief_research_packet
ChatGPTSelected-skill-only research packet for news-brief. Normalize the user request into one typed market, asset, or shared-item intent, then call exactly once. Asset mode accepts one or more model-classified subjects; item mode accepts a web URL, X URL, or pasted text. It owns identity verification, event windows, structured provider fan-out, deduplication, deterministic importance scoring, reaction evidence, typed gaps, and resumable clarification repair. Exact web-page reading This call consumes 25 StableJack credits before execution, including when provider execution later fails.
perpetual_asset_screener
ChatGPTScreens, ranks, and filters Hyperliquid perp assets across default perps and HIP-3 builder dex universes. Use this tool for broad market scanning questions like "top 10 assets by volume", "bottom 20 by funding rate", "assets on xyz with volume > 1M", or "show me the highest OI assets across default perps and flx". Supports flexible boolean filter expressions (AND/OR) and deterministic ranking by volume24hUsd, openInterestUsd, fundingRate, markPrice, change24hPct, or premium. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
perpetual_market_context
ChatGPTFetches live perp market context for a single symbol from Hyperliquid. Returns pricing (mark, mid, oracle), activity (24h volume, OI), funding rate, and market metadata (leverage, decimals). The markPrice field is the AUTHORITATIVE current price for this asset — use it when the user asks for the "current price" or "price right now". Supports HIP-3 builder DEX scoping via the optional dex parameter or namespace syntax like xyz:CL. Common commodity/equity names are auto-resolve This call consumes 25 StableJack credits before execution, including when provider execution later fails.
post_earnings_research_packet
ChatGPTSelected-skill-only research packet for post-earnings-result-review. Call once with a canonical listed-company symbol, or with a raw subject plus model-authored assetIntent. Use assetIntent=equity only for a listed operating company; non-equity intents return typed scope repair. It owns issuer resolution, scope validation, provider reads, retries and typed repair output; it never requires the model to call an internal discovery or provider tool. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
predicted_funding_rates
ChatGPTRetrieves predicted funding rates for Hyperliquid perp assets across multiple exchanges (Binance, Bybit, Hyperliquid, etc.). Returns per-asset predictions with funding rate, next funding time, and interval. Optionally filter by a list of coin symbols; omit for all assets. Note: only supports the default (first) perp universe — HIP-3 builder dex assets are not included. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
protocol_events
ChatGPTDefiLlama dated hallmarks for one explicitly named DeFi protocol only, such as a launch, migration, incentive change, or protocol milestone. Do not use as a news search, or for hacks, raises, chain events, market catalysts, or broad event discovery. --- Native skill: defi-data --- --- name: defi-data description: Core reference for DefiLlama MCP tools. Maps DeFi questions to the correct tool call with proper parameters. Covers entity conventions, metric interpretation, stoc This call consumes 25 StableJack credits before execution, including when provider execution later fails.
protocol_income_statement
ChatGPTDefiLlama on-chain DeFi PROTOCOL financial-statement breakdown only: fees, protocol revenue, incentives, and token-holder distributions by month, quarter, or year. This is not a public-company income statement. Never use it for stocks or corporate financials; use FMP equity.financial-statements for those. --- Native skill: defi-data --- --- name: defi-data description: Core reference for DefiLlama MCP tools. Maps DeFi questions to the correct tool call with proper parameter This call consumes 25 StableJack credits before execution, including when provider execution later fails.
protocol_intelligence
ChatGPTReturns a combined protocol-level analytics snapshot from DefiLlama in a SINGLE call. Always fetches protocol info (TVL, chain breakdown, metadata, 7d trend, hallmarks) AND metrics (fees, revenue, DEX volume, derivatives, options) together. Optionally attaches the FULL daily TVL history envelope at info.tvlHistory (series + window + rangeStats + full-history stats incl. ATH/ATL/drawdown/momentum) when includeTvlHistory is true OR any of fromDate / toDate / lastN is This call consumes 25 StableJack credits before execution, including when provider execution later fails.
social_market_research
ChatGPTSearches X posts, user profiles, and threads via xAI X Search. Use this for current X/Twitter social sentiment, breaking narrative checks, account/thread lookup, and X-native source discovery. For general web/news search, use Exa instead. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
stock_insight_by_symbol
ChatGPTGet the current StableJack stock insight record for one Hyperliquid XYZ equity-perp symbol. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
stock_insights
ChatGPTGet all current StableJack stock insight records for the Hyperliquid XYZ equity-perp universe. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
technical_indicator_catalog
ChatGPTLists the TA-Lib indicator/function catalog without bloating the main indicator tool description. Use this when you need to discover available TA-Lib functions, groups, required inputs, optional params, and whether JackIndicator can compute them from Hyperliquid single-symbol OHLCV candles. Use entry.name/code for jack-indicator tool calls. Use entry.label only for user-facing display. For user-facing answers, always show entry.label or "CODE - Full descriptive name"; do not This call consumes 25 StableJack credits before execution, including when provider execution later fails.
token_intelligence
ChatGPTReturns a combined token-level analytics snapshot from DefiLlama in a SINGLE call. Provide symbol (e.g. "HYPE", "ETH") to get TVL exposure (which protocols hold the token). Provide protocol slug (e.g. "hyperliquid", "aave") to get emission/unlock schedule. Provide BOTH for the fullest picture — both are fetched in parallel. At least one of symbol or protocol must be given. Endpoints: /api/tokenProtocols/{symbol}, /api/emission/{protocol}. totalProtocols and `totalEx This call consumes 25 StableJack credits before execution, including when provider execution later fails.
token_unlocks
ChatGPTDefiLlama crypto-token vesting and dated unlock schedules only, including cliff or linear allocations and token supply context. Use only when the user asks when/how much a token unlocks. Do not use for live token price, protocol TVL, generic token research, or public-company lockups. --- Native skill: defi-data --- --- name: defi-data description: Core reference for DefiLlama MCP tools. Maps DeFi questions to the correct tool call with proper parameters. Covers entity conve This call consumes 25 StableJack credits before execution, including when provider execution later fails.
top_asset_positions
ChatGPTFetches the largest / most profitable open positions on a specific coin from Hyperliquid crowd data. Use this for "who are the top traders on OIL?", "biggest BTC positions", "largest whale positions on ETH", "who is making money on SOL?". Supports filtering by side (long/short) and crowd segment (whales, smart money, etc.). Omit dex to hit the canonical listing (OIL→xyz:CL, GOLD→xyz:GOLD, TSLA→xyz:TSLA). Only set dex when the user explicitly names one or asks to compare l This call consumes 25 StableJack credits before execution, including when provider execution later fails.
top_trader_intelligence
ChatGPTFetches the top (or bottom) traders on Hyperliquid by PnL or ROI over a chosen time window. Use this for "who is making money?", "top traders this week", "best ROI this month", "worst performers today". --- Native skill: SKILL.md --- HyperTracker API Pre-computed intelligence layer for Hyperliquid, the largest on-chain perpetual futures exchange. Cohort analytics, live order flow, liquidation risk scoring, leaderboards, and more. All data is independently aggregated and This call consumes 25 StableJack credits before execution, including when provider execution later fails.
valuation_research_packet
ChatGPTSelected-skill-only research packet for valuation-check. Call once with a canonical listed-company symbol, or with a raw subject plus model-authored assetIntent. Use assetIntent=equity only for a listed operating company; non-equity intents return typed scope repair. It owns issuer resolution, scope validation, provider reads, retries and typed repair output; it never requires the model to call an internal discovery or provider tool. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
web_page_research
ChatGPTexa connector tool exa_web_fetch_exa; capabilities: market.news, market.macro, crypto.news. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
web_research
ChatGPTStandard simple Exa public-web search fallback for public facts that authorized structured and native data tools cannot sufficiently cover. Use only after higher-priority sources return typed empty, omit the requested field or period, return the wrong scope, or leave materially incomplete evidence. Search public web pages and news for the missing evidence before concluding that the data is unavailable. Capabilities: market.news, market.macro, crypto.news. This call consumes 25 StableJack credits before execution, including when provider execution later fails.
