> For the complete documentation index, see [llms.txt](https://rootai.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://rootai.gitbook.io/docs/mcp.md). # MCP ## What is the rootAI MCP Server? Model Context Protocol (MCP) is an open standard that lets AI assistants call external tools and live data sources in real time instead of relying on stale training data. The rootAI MCP server sits on top of Edge's data layer, giving any MCP-compatible AI assistant live access to crypto market intelligence: prices, funding rates, open interest, orderbook depth, options chains, news, on-chain discovery, Base swaps, TradFi stocks and more — across Hyperliquid, Binance, Pyth, DexScreener and Base. ``` Your AI assistant → rootAI MCP server → Edge data layer → live exchange + on-chain data ``` **Public endpoint:** `https://mcp.rootedge.ai/mcp` The endpoint is read-only and non-custodial. Market data, news, signals, discovery, and unsigned Base transaction calldata. The MCP never holds keys; on-chain actions are returned as unsigned calldata for the user to sign in their own wallet. ### Setup Guides #### Claude.ai: Web (Recommended) No installs. Under a minute. 1. Go to **claude.ai** and sign in 2. Navigate to **Customize → Connectors**
1. Click **Add custom connector**
1. Paste the URL: `https://mcp.rootedge.ai/mcp`
1. Verify by asking: > What perp markets are live on Hyperliquid right now? A list of trading pairs means you're connected.
#### Claude Desktop 1. Open or create the config file * **macOS:** `~/Library/Application Support/Claude/claude_desktop_config.json` * **Windows:** `%APPDATA%\Claude\claude_desktop_config.json` 2. Add the rootAI server: ```json { "mcpServers": { "rootai": { "command": "npx", "args": ["-y", "@modelcontextprotocol/server-fetch"], "env": { "MCP_SERVER_URL": "https://mcp.rootedge.ai/mcp" } } } } ``` 3. Restart Claude Desktop fully (not just the window) 4. Look for the 🔌 tools indicator in the chat input — confirms MCP is active #### Cursor 1. Open **Settings → MCP** 2. Add a new server entry: ```json { "rootai": { "url": "https://mcp.rootedge.ai/mcp", "type": "sse" } } ``` 3. Reload the window and the server will appear in your MCP tools list ## Complete Tool Reference ### `hyperliquid_*` — Hyperliquid Data * `hyperliquid_meta` — universe of perpetuals, decimals, max leverage * `hyperliquid_summaries` — 24h volume, open interest, mark price * `hyperliquid_mids` — best mid for every coin * `hyperliquid_book` — full L2 orderbook * `hyperliquid_candles` — OHLCV candles by interval * `hyperliquid_funding` — current + historical funding rates * `hyperliquid_stock_rwa` — HL-listed RWA / equity perps ### `binance_public` Binance public market data (spot / USD-M / Coin-M). ### `dex_*` — DexScreener (token & pair discovery) * `dex_search` — search tokens / pairs by symbol, name, or contract address * `dex_pair` — full market data for a single pair * `dex_tokens`, `dex_token_pairs`, `dex_token_orders` — token-level lookups * `dex_trending_metas`, `dex_meta` — narrative discovery * `dex_top_boosts`, `dex_latest_boosts`, `dex_latest_ads` — paid promo signals * `dex_latest_profiles`, `dex_recent_profile_updates`, `dex_latest_community_takeovers` #### `polymarket_*` — Polymarket Prediction Markets For any specific event question, always start with `polymarket_search` using the most precise query you can build (include sport, league, year). Avoid `polymarket_events` as a first move, it returns an unfiltered listing and is easy to pick the wrong event from. If the user's term is ambiguous (multiple sports or seasons share the name), enrich the query first or ask. Always check each candidate event's **status** before quoting odds, closed/resolved events return last-known prices that are not live. **Discovery** * `polymarket_search` — full-text search across markets, events, and series * `polymarket_events` — list events (browsing only; use for "what's trending") * `polymarket_event`, `polymarket_event_by_slug` — single event with all its child markets * `polymarket_event_tags`, `polymarket_market_tags`, `polymarket_tags`, `polymarket_tag` — taxonomy * `polymarket_series`, `polymarket_series_detail` — multi-event series (e.g. NFL season) * `polymarket_sports`, `polymarket_teams` — sports league + team metadata **Markets** * `polymarket_markets` — list markets with filters (active, closed, by tag, by event) * `polymarket_market`, `polymarket_market_by_slug`, `polymarket_market_by_token` — single-market lookup * `polymarket_simplified_markets` — compact market list for quick scans * `polymarket_clob_markets` — CLOB-tradeable markets only **Prices & quotes** * `polymarket_price`, `polymarket_prices` — last trade price for one / many tokens * `polymarket_last_trade_price`, `polymarket_last_trade_prices` — most recent fill * `polymarket_midpoint`, `polymarket_midpoints` — mid between bid/ask * `polymarket_spread`, `polymarket_spreads` — bid/ask spread * `polymarket_tick_size` — minimum price increment for a market * `polymarket_prices_history` — historical price series for charting **Orderbook & flow** * `polymarket_orderbook`, `polymarket_orderbooks` — L2 book for one / many tokens * `polymarket_trades` — recent trade tape * `polymarket_open_interest` — outstanding YES/NO contracts per market * `polymarket_holders` — wallet-level holder distribution **Wallet activity (public onchain data)** * `polymarket_activity` — recent activity for a wallet * `polymarket_positions`, `polymarket_closed_positions` — open + realized positions * `polymarket_user_value` — total portfolio value for a wallet ### `base_*` — Base Chain (calldata) * `base_prepare_swap` — swap calldata for the user's wallet * `base_prepare_send` — ERC-20 / ETH transfer calldata These tools never broadcast. They return the unsigned `calls` array for the user to approve and sign in their own wallet. ### `stock_market_cissor_*` — TradFi & Stocks * `stock_market_cissor_search`, `stock_market_cissor_quote`, `stock_market_cissor_multi_quote` * `stock_market_cissor_history_daily`, `stock_market_cissor_history_intraday` * `stock_market_cissor_market_summary`, `stock_market_cissor_sector_data`, `stock_market_cissor_industry_data` * `stock_market_cissor_fundamentals_summary`, `stock_market_cissor_balance_sheet`, `stock_market_cissor_cash_flow`, `stock_market_cissor_income_statement` * `stock_market_cissor_earnings_dates`, `stock_market_cissor_corporate_actions`, `stock_market_cissor_ticker_news` * `stock_market_cissor_options_chain`, `stock_market_cissor_options_summary`, `stock_market_cissor_options_expirations` ### `edge_*` — Edge Intelligence Layer * `edge_news_latest`, `edge_news_by_asset` — aggregated crypto news * `edge_fear_greed` — composite 0–100 sentiment regime * `edge_pyth_price`, `edge_pyth_feeds` — Pyth oracle prices * `edge_bankr_launches`, `edge_discover_base_token_opportunities` — Base ecosystem discovery ### Fear & Greed Index `edge_fear_greed` returns a composite 0–100 score built entirely from on-platform data — BTC volatility, momentum, market-wide funding extremity, BTC open-interest shift, and news tone. The score is **directional, not predictive.** It reflects current market conditions from on-platform data only and uses no third-party sentiment APIs. ### Practical Examples **Morning Market Brief** ``` Give me a morning brief: - BTC and ETH prices and 24h change on Hyperliquid - Top 3 movers on Hyperliquid - Current Fear & Greed score and what it implies - Any FUNDING_EXTREME or MEGA_TRADE signals from the last hour - Latest market news headlines ``` **Pre-Trade Research** ``` I'm thinking of longing SOL on Hyperliquid. Can you: 1. Check the current price and 24h trend 2. Show me the funding rate — am I paying or getting paid to hold long? 3. Check recent OI shifts on SOL-USD-PERP 4. Pull any SOL news from the last 24 hours 5. Check the orderbook depth for a $20k position — how much slippage should I expect? ``` **Base Discovery** ``` Find the most promising Base launches from the last 24 hours. Score them by liquidity, volume, holder count, and recent boosts. For the top 3, pull DexScreener pair data and any news mentions. ``` **Polymarket Scan** ``` Scan Polymarket for high-conviction setups: 1. Pull the top 10 most active events by volume right now 2. For each, show the leading market's price, midpoint, and spread 3. Flag any market where the orderbook shows >$50k depth on one side 4. Cross-check the implied probability against any recent news on the topic 5. Surface mispricings where on-chain flow (holders, open interest) diverges from price ``` **Base Swap** ``` Prepare a swap of 0.01 ETH to USDC on Base for wallet 0x4qfa…560314 with 0.5% slippage. Return the unsigned calldata so I can sign it in my wallet. ``` If you want to execute on Base directly from chat, use the Edge dashboard at . It ingests the MCP's `base_prepare_*` calldata and forwards it to your connected wallet for signing. ### Troubleshooting Still stuck? File an issue with: * The tool name + arguments you called * The exact error string * Your MCP client (claude.ai web / Claude Desktop / Cursor / other) + version ### Tips for Best Results * **Be explicit about the venue.** Say *on Hyperliquid*, *on Base*, or *on Polymarket,* some tools require it. * **Chain tools in one prompt.** MCP calls stack: ask for price + funding + signals + news in a single message for a full picture. * **Ask for interpretation, not just data.** *Is this funding rate extreme? What does it imply for positioning?* gets you more than a raw number. * **Fear & Greed is directional, not a trade signal.** Use it for regime context alongside price and funding data. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://rootai.gitbook.io/docs/mcp.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.