Heisenbergdocs

AI Agents

Give your AI agents live prediction market intelligence through Claude MCP or direct API calls.

Claude MCP Integration

The fastest way to add prediction market awareness to an AI agent. Connect Heisenberg as an MCP server and Claude can query markets, analyze wallets, and surface opportunities in natural language.

Setup guide: See Work with AI for step-by-step MCP setup.

Example agent prompts:

  • •"What are the highest-volume open markets on Polymarket?"
  • •"Analyze wallet 0x1234... — is this a good trader to follow?"
  • •"Are there any sharp price moves on the US election market in the last hour?"
  • •"Compare Polymarket and Kalshi odds for the Fed rate decision"

Autonomous Trading Agents

Build agents that autonomously monitor and trade prediction markets:

  1. 1Monitor: Poll Social Pulse for breaking news and Price Jumps for anomalies.
  2. 2Analyze: Use Market 360 and Orderbook to assess market conditions.
  3. 3Validate: Cross-reference with smart money positions via Trades to see how top wallets are positioned.
  4. 4Execute: Place trades based on your agent's strategy and risk parameters.

RAG & Knowledge Retrieval

Use Heisenberg as a live data source for Retrieval-Augmented Generation. The structured JSON responses are designed to be easily ingested by LLMs. Pass API results directly into your agent's context window for grounded, fact-based responses about prediction markets.