PolyDesk
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@PolyDesk search for markets about AI regulation"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
PolyDesk — Polymarket AI Research & Trading Control Plane
An MCP server that turns Polymarket prediction markets into a structured research and trading workflow — powered by Notion as the dashboard and knowledge base. Zero-interaction automation: check a box in Notion, AI research appears automatically.
Inspired by Karpathy's autoresearch: the AI agent runs iterative research loops on prediction markets, scores them, and writes structured findings to Notion for human review and trade execution.
Screenshots
Market Watchlist (left columns)

Market Watchlist (right columns) — Research checkbox triggers AI analysis

Research Reports Database

One-Command Setup via Claude Desktop

MCP Server Integration

Related MCP server: SimpleFunctions
Architecture
┌──────────────┐ ┌──────────────────┐ ┌──────────────┐
│ Polymarket │────→│ PolyDesk MCP │────→│ AI Agent │
│ Gamma API │ │ (this server) │ │ (Claude, etc)│
└──────────────┘ └──────────────────┘ └──────┬───────┘
│
▼
┌──────────────┐
│ Notion MCP │
│ (official) │
└──────┬───────┘
│
▼
┌──────────────┐
│ Notion │
│ Workspace │
│ (Dashboard) │
└──────────────┘PolyDesk MCP provides Polymarket data, AI research (via local Ollama), and direct Notion integration. Notion MCP (official) provides additional Notion read/write capabilities. Together they create a zero-interaction trading control plane — all from Notion.
Key Features
One-command setup —
call polydesk setupcreates the full Notion workspace with databases and trending marketsZero-interaction research — Check the 🔬 Research box in Notion, AI analysis appears automatically (10s polling)
Keyword search — Type a keyword (e.g. "Trump") in a new row, the system finds the matching Polymarket market
Local AI — Research powered by Ollama (qwen2.5:14b) running locally, no cloud API needed
Human-in-the-loop — AI generates signals, humans approve trades via Notion
Tools (22 total)
Tool | Description |
| Create Notion workspace with databases and trending markets |
| Discover hottest markets by volume |
| Search markets by keyword |
| Full details for a specific market |
| Browse top events (grouped markets) |
| Details for a specific event |
| Bulk price feed for multiple markets |
| Structured research prompt (autoresearch pattern) |
| Run local AI research on a market |
| Research multiple markets in one loop |
| Position sizing, R:R, Kelly criterion |
| Live P&L for open positions |
| Side-by-side odds comparison |
| Heuristic mispricing detector |
| Generate rich Notion blocks from research |
| Format market data for Notion watchlist |
| Sync watchlist with Polymarket prices |
| Start background polling for Notion changes |
| Stop background polling |
| Check for human trade approvals |
| Create trade execution plan |
Resources
Resource | Description |
| Notion database schemas for Watchlist and Research |
Prompts
Prompt | Description |
| Bootstrap the full Notion workspace |
| Run a complete research cycle on trending markets |
| Sync positions and generate P&L summary |
How It Works
Setup —
call polydesk setupcreates a Notion page with Watchlist and Research Reports databases, populated with live trending marketsAuto-watch — Background polling starts automatically, checking Notion every 10 seconds
Research trigger — Check the 🔬 Research box on any market row → status changes to "Researching" → Ollama runs analysis → research report appears in the Research Reports database
Keyword search — Add a new row with just a keyword → system searches Polymarket and fills in market data
Human review — Review AI signals and set Human Approval to "Approved" → system generates execution plan
Iterate — Re-check the research box to run deeper analysis iterations (ratchet mechanism)
Setup
Prerequisites
Ollama with
qwen2.5:14bmodel (ollama pull qwen2.5:14b)
1. Install
npm install
npm run build2. Configure Claude Desktop
Add both MCP servers to your claude_desktop_config.json:
{
"mcpServers": {
"polydesk": {
"command": "node",
"args": ["/path/to/polydesk-mcp/dist/index.js"],
"env": {
"NOTION_TOKEN": "ntn_YOUR_TOKEN"
}
},
"notion": {
"command": "npx",
"args": ["-y", "@notionhq/notion-mcp-server"],
"env": {
"OPENAPI_MCP_HEADERS": "{\"Authorization\": \"Bearer ntn_YOUR_TOKEN\", \"Notion-Version\": \"2022-06-28\"}"
}
}
}
}3. Bootstrap Your Workspace
In Claude Desktop, say: "Call polydesk setup"
This creates the full Notion workspace and starts auto-watching.
4. Start Researching
Just check the 🔬 Research box on any market in Notion. No commands needed — research appears automatically.
You can also use Claude Desktop directly:
"Scan trending markets and research the top 5"
"Find markets about bitcoin and analyze the best opportunity"
"Run the daily research loop"No Polymarket API Key Required
PolyDesk uses the public Polymarket Gamma API — no authentication needed for market data.
Development
npm run dev # Run with tsx (hot reload)
npm run check # Type-check
npm run build # Compile to dist/License
MIT
Available Tools
21 toolsauto_research_marketB
Generate a structured research prompt for a Polymarket market. Inspired by Karpathy's autoresearch — the AI agent should analyze the prompt, form a thesis, then write findings to Notion via the Notion MCP server. Supports iterative ratchet: re-run with higher iteration to refine.
Inputs: market_id (required), iteration.
batch_researchB
Generate research prompts for multiple trending markets at once. The agent should process each, analyze, and write reports to Notion.
Inputs: limit, min_volume_usd.
calculate_tradeB
Calculate trade parameters: position sizing, risk/reward ratio, max gain, max loss. Use before logging a trade.
Inputs: side (required), size_usd (required), market_id (required), stop_loss (required), +2 more.
check_positionsA
Fetch current prices for a list of market IDs and calculate P&L against entry prices. Use this to update the Notion Trade Journal.
Inputs: positions (required).
compare_marketsB
Compare odds across multiple markets side by side. Useful for finding relative value or arbitrage opportunities.
Inputs: market_ids (required).
edge_scannerA
Scan trending markets and flag those where simple heuristics suggest potential mispricing. A quick filter before deep research.
Inputs: limit.
format_research_for_notionA
Generate beautiful, rich Notion page content blocks for a market research report. Returns the exact Notion API block structure — pass it directly to the Notion MCP's API-patch-block-children tool to create a stunning research page.
Workflow:
First create a page in your Research database using API-post-page
Then call this tool to get the formatted blocks
Pass the blocks to API-patch-block-children with the page ID
Inputs: risks (required), analysis (required), evidence (required), base_rate (required), +5 more.
format_watchlist_entryA
Generate the properties object for adding a market to the Notion Watchlist database. Returns the exact Notion API properties structure — pass it to the Notion MCP's API-post-page tool.
Inputs: market_id (required), category, conviction, fair_value.
generate_execution_planA
Given a list of human-approved markets with fair values, generate a portfolio execution plan with position sizing (Kelly criterion), risk limits, and correlation checks. Use after reading human approvals from Notion.
Inputs: approved_markets (required), total_bankroll, max_portfolio_risk.
get_eventA
Get full details for a Polymarket event and all its markets.
Inputs: event_id (required).
get_eventsB
Fetch top Polymarket events (event = group of related markets). Great for finding multi-market opportunities.
Inputs: limit.
get_marketA
Get full details for a Polymarket market: description, outcomes, prices, volume, liquidity.
Inputs: market_id (required).
get_pricesA
Get current prices for one or more markets by ID. Use this to refresh watchlist or check positions.
Inputs: market_ids (required).
research_with_ollamaA
One-click AI research: fetches market data, calls local Ollama LLM for analysis, and returns both the research results AND formatted Notion blocks ready to write. This is the "Research Button" — call it, then write the output to Notion.
Returns:
Research results (conviction, fair value, edge, evidence, risks, analysis)
Notion blocks for a beautiful research report page (pass to API-patch-block-children)
Watchlist properties (pass to API-post-page or API-patch-page)
Requires Ollama running locally (default: http://localhost:11434).
Inputs: market_id (required), iteration.
scan_trending_marketsA
Fetch the hottest active Polymarket markets ranked by volume. Use this to discover trading opportunities.
Inputs: limit.
search_marketsA
Search Polymarket for markets matching a keyword (e.g., 'bitcoin', 'trump', 'fed', 'election').
Inputs: query (required), limit.
setupA
Set up the PolyDesk trading desk in Notion. Creates databases, populates with live Polymarket data, and starts auto-watching. Call when user says "setup", "set up", "create trading desk", or "start". Does everything directly — no extra steps needed.
Inputs: market_count.
sync_notion_watchlistA
Bidirectional sync between Notion Watchlist and Polymarket/Ollama. Call this when user says "sync". Reads your Notion Watchlist database and processes changes:
🔬 Research checkbox: If checked (true), runs Ollama AI research, writes report, then UNCHECKS the box
Keyword → Market: If a row has a keyword in the Market column but no Market ID, searches Polymarket and populates all data
Approval action: If Human Approval = "Approved", generates an execution plan
Re-research: If Human Approval = "Needs More Research", runs research again with iteration+1
Price refresh: Updates all prices to latest
Example: User types "Trump" in Notion → runs sync → system finds best matching Polymarket market and fills in all data.
Inputs: watchlist_database_id (required), run_research, research_database_id.
unwatchA
Stop auto-watching the Notion Watchlist. Stops the polling loop started by watch.
Inputs: no named parameters.
validate_human_overridesA
Check if human fair value overrides are reasonable by comparing against market data and flagging potential errors. Use before executing approved trades.
Inputs: overrides (required).
watchA
Start auto-watching the Notion Watchlist. Polls every N seconds for checkbox changes and keyword entries. When user checks 🔬 Research in Notion, AI research runs automatically — no need to say "sync".
Requires NOTION_TOKEN env var. Call "unwatch" to stop.
Inputs: watchlist_database_id (required), interval_seconds, research_database_id.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 21 tools
Most market-data tools are clearly separated by resource, but the research tools overlap: auto_research_market, research_with_ollama, and batch_research all appear to initiate research, and setup vs watch both start polling. Descriptions help, but an agent could easily invoke the wrong research workflow.
The majority follow verb_noun snake_case (get_market, search_markets, calculate_trade, generate_execution_plan). Deviations like edge_scanner, setup/unwatch, batch_research, and research_with_ollama break the pattern, but the conventions remain readable and mostly predictable.
Twenty-one tools is substantial, and the breadth requires several one-off helpers (format_watchlist_entry, format_research_for_notion, unwatch). Most tools earn their place for a Polymarket+Notion+Ollama workstation, but the surface feels heavy and could be consolidated.
The core loop is covered: discover markets, research, sync with Notion, generate execution plans, and check positions. However, there is no tool to actually log trades or place orders, despite calculate_trade/check_positions referring to a trade journal workflow, and no direct way to view or delete watchlist entries beyond sync.
Related MCP Connectors
Polymarket + Hyperliquid + macro for AI agents. 38 tools, signal backtest, SSE streaming. Free tier.
AI prediction market trading — real-time data, CLOB exchange, Prop Desk, 40+ tools
Polymarket in your AI: market rules, real fill costs, who is trading, top traders. Read-only.
Calibrated world model for AI agents. 40 tools: world state, markets, trading. Kalshi + Polymarket.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables Claude to autonomously trade, analyze, and manage positions on Polymarket prediction markets with 45 comprehensive tools covering market discovery, analysis, trading execution, portfolio management, and real-time monitoring with enterprise-grade safety features.MIT
- AlicenseAqualityBmaintenance24/7 autonomous monitoring and edge detection for prediction markets (Kalshi & Polymarket). Features causal tree analysis, orderbook depth tracking, cross-venue comparison, and real-time alerts.16196 npm12MIT
- AlicenseAqualityDmaintenanceProvides prediction market intelligence, research, and strategy signals for platforms like Kalshi, Polymarket, and Robinhood. It enables AI assistants to perform market screening, arbitrage detection, and deep causal analysis to support informed trading decisions.271MIT
- AlicenseAqualityDmaintenanceTrade, analyze, and automate Polymarket prediction markets via AI. 34 tools for direct trading, smart money flow, copy trading, backtest, and portfolio management.48115 npm20MIT