Report Abuse

Basic Information

Polymarket Agents is a developer-focused framework and utilities repository for building AI agents that interact with Polymarket prediction markets. The code provides connectors and clients to fetch market and event metadata, sign and execute orders on Polymarket, and run autonomous trading logic. It is intended for Python 3.9 and includes command line tools and example scripts to run agents locally or in Docker. The project supports prompt engineering and LLM integration and uses environment variables for API keys and a polygon wallet private key. It also includes guidance for loading a wallet with USDC before trading. The project is distributed under the MIT License and is designed for developers who want to prototype, extend, or deploy AI-driven trading agents that use external data and vectorized context.

Links

App Details

Features
The repository includes a set of core modules and utilities: connectors for Polymarket API, a Gamma market client to parse market and event metadata, a Polymarket client to retrieve markets and execute trades, and Pydantic data models in Objects.py. It provides local and remote retrieval-augmented generation support and a Chroma-based example for vectorizing news and API data. A CLI (scripts/python/cli.py) exposes commands like get-all-markets and other utilities to query data, call LLMs, and execute trades. It supplies example application scripts such as agents/application/trade.py, Docker build and run scripts, environment variable examples, and instructions for dependency installation and pre-commit hooks.
Use Cases
This project standardizes common agent development tasks for prediction markets so developers can focus on strategy rather than infrastructure. It centralizes market data access, order-building and signing utilities, and models for trades and events, enabling reproducible agent behavior. Retrieval-augmented generation and vector database support let agents incorporate news and external context into decisions. The CLI and example trade script allow quick experimentation and manual control, while Docker and environment guidance simplify deployment and testing. The repo lowers the barrier to integrating LLMs, external data sources, and automated trading on Polymarket, while the MIT license and contributor guidance facilitate community extension and reuse.

Please fill the required fields*