Practical overview
When it fits
Use it when developing or evaluating a Python prediction-market agent and you want benchmark reports or the documented deployment and monitoring interfaces. The README recommends using your own infrastructure for deployment rather than its deprecated deployment approach.
Who it is for
- Python developers building agents for prediction-market applications.
- Teams comparing agent predictions against crowd judgments.
- Developers working with the listed Manifold, AIOmen or Polymarket integrations.
Documented uses
- Benchmark an agent's predictions against crowd judgments on selected markets.
- Generate a Markdown report to compare agents side by side.
- Build a deployable agent using the documented agent interface.
- Monitor GCP-deployed agents and prediction-market platform metrics through a local Streamlit app.
- Inspect market abstractions for retrieving markets and buying or selling tokens.
Requirements & limits
The repository is public and specifies the LGPL-3.0 license. The project requires Python 3.10 or later and uses Poetry; GCP deployment and monitoring require the gcloud CLI and authorization.
- Platform support differs by task: Manifold and AIOmen are listed for benchmarking, deployment and monitoring; Polymarket is listed for benchmarking only.
- The README marks its deployment approach deprecated and recommends using your own infrastructure.
- The market-actions Python API is described as currently undocumented, with the AgentMarket class suggested for inspection.
- The setup example lists Manifold, betting private-key and OpenAI API key environment variables.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up the Python project environment
Install the project dependencies and open its Poetry shell.
Before you start
- Python 3.10 or later
- Poetry
Steps
Install Poetry with the command below.
python3.10 -m pip install poetryFrom the repository root, run the command below.
python3.10 -m poetry installEnter the environment with the command below.
python3.10 -m poetry shell
Expected outcome
The project dependencies are installed and the Poetry shell is active.
Guide limitations
- These steps cover environment setup only, not running an agent or benchmark.
Benchmark an agent on binary Manifold markets
Run the documented sample benchmark and generate a Markdown comparison report.
Before you start
- Python 3.10 or later
- Project dependencies installed with Poetry
- An agent implementing the benchmark interface; the README example uses the included RandomAgent
- Markets available from the selected prediction-market platform
Steps
Create a benchmarkable agent by subclassing
AbstractBenchmarkedAgentand connecting its research and prediction functions inpredict, or use the documentedRandomAgentexample.Create a
Benchmarkerwith binary markets from Manifold and the agent or agents to compare.Run
benchmarker.run_agents().Generate the report with
benchmarker.generate_markdown_report().
Expected outcome
A Markdown report for comparing agent predictions with the crowd on the selected markets.
Guide limitations
- The example uses binary Manifold markets and does not document a general benchmark configuration for every platform.
- The guide describes the README sample workflow and is not a claim of execution or verification.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 14 October 2026
- Original website title
- gnosis/prediction-market-agent-tooling · Official homepage title · retained in its source language