Simmer SDK: automation for Hermes Agent

Prediction market trading skills, MCP server and Python SDK for AI agents on Polymarket and Kalshi

What Simmer SDK does

Simmer SDK is the open-source part of Simmer, a platform where AI agents trade prediction markets through one unified API. The repository holds three things: 28 skills an agent installs and runs, an MCP server published to npm as simmer-mcp, and a Python SDK published to PyPI as simmer-sdk. The venues named in the README are Polymarket, Kalshi and Hyperliquid, plus a sim venue that paper-trades with virtual $SIM currency.

The SDK client lists markets, returns trading context such as safeguards, slippage and conflict detection, places trades, and reports positions and profit and loss. Live execution is a constructor argument, so the default pattern previews paper trades and only places real orders when a skill is run with --live. Configuration comes from environment variables: SIMMER_API_KEY, TRADING_VENUE and, for Polymarket self-custody, WALLET_PRIVATE_KEY. The README's main quick start follows the OpenClaw skill pattern, and the repository is also tagged hermes-agent.

Key features

28 installable trading skills for agents
MCP server published to npm as simmer-mcp
Python SDK published to PyPI as simmer-sdk
Unified API across Polymarket, Kalshi and Hyperliquid
Paper trading with simulated $SIM currency before real funds
Market context with safeguards, slippage and conflict detection

When to use it

Letting an agent paper-trade prediction markets and review its profit and loss
Building a custom trading skill on top of the Python SDK
Connecting an MCP-capable agent to prediction market data and orders

Who it is for: Developers and agent users who want to run autonomous prediction market trading strategies with safety rails.

How it fits with Hermes Agent

Simmer works with agents that install skills or use MCP. The repository is tagged hermes-agent and openclaw, and the README quick start is written for OpenClaw.

How to install Simmer SDK

These commands are copied from the project's README. Check the repository for the latest steps before you run them.

pip install simmer-sdk

Requirements: A Simmer API key from simmer.markets/dashboard. Polymarket self-custody trading also needs a wallet private key.

Note: It can place real-money trades, so the README advises dry-run and paper trading with $SIM before going live, and it requires a Simmer API key.

FAQ

What is Simmer SDK?

Simmer SDK is a toolkit for letting AI agents trade prediction markets. It bundles installable skills, an MCP server and a Python SDK for the Simmer platform.

Does Simmer SDK work with Hermes Agent?

The repository is tagged hermes-agent and ships skills and an MCP server that compatible agents can use. Its README quick start, however, follows the OpenClaw skill pattern.

How do I install Simmer SDK?

Run pip install simmer-sdk and get an API key from simmer.markets/dashboard. Then set SIMMER_API_KEY and TRADING_VENUE as environment variables.