A research agent tool that takes an LLM prompt and returns generated text to identify and propose new prediction topics.

Creates a prediction market with a defined title and set of outcomes, returning the deployed market address and the transaction hash.

Adds a new outcome option to an existing prediction market by submitting an on-chain transaction and returning the resulting transaction hash.

Searches and retrieves relevant user memories or stored facts based on query terms, optionally returning full documents with configurable result limits.

Returns the total number of prediction markets deployed by the factory.

Reads on-chain market metadata by index, returning the market’s contract address, title, outcome token details (supply and price), and settlement status.

Executes a market position by approving collateral if required and purchasing a specified number of Yes or No shares for a given outcome, returning transaction details and cost.

Checks whether a market image blob already exists in Vercel Blob storage for a given market address.

Generates a prompt-based image for a prediction market, stores it in Vercel Blob storage, and returns the image URL.

Searches X for topic-matched tweets and returns candidate users to engage with, including their tweet content, author, and tweet URL.

Posts a reply tweet to a specified X (Twitter) status, contextualized by a given prediction market and the original tweet content.

Allows the agent to perform keyword search, semantic search, user search, and thread fetch on X

Allows the agent to search the web and browse pages