# openai/universe-starter-agent

A starter agent that can solve a number of universe environments.

Repository: https://github.com/openai/universe-starter-agent
Canonical: https://ross.abutalabs.com/products/universe-starter-agent
Language: Python
License: MIT
License Family: permissive
Archived: true
Last push: 2018-04-07T02:30:10+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3576, "days_push": 3071, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1097, forks 313 (observed 2026-08-28T04:03:34.691977+00:00)

## What it is
A starter implementation of an A3C reinforcement learning agent that can solve OpenAI Universe environments, including Atari games like Pong. It is a reference codebase for learning policy-gradient RL, now deprecated in favor of OpenAI's Retro library.

## Use cases
- train an A3C agent on Atari Pong
- learn how asynchronous policy gradient RL is implemented
- get started with OpenAI Universe environments
- run parallel RL workers with a parameter server
- study a real-time RL agent codebase in TensorFlow

## When to choose
- you want a minimal, readable A3C reference implementation
- you are learning reinforcement learning with TensorFlow and OpenAI Gym
- you need a starting point for agents on Universe/VNC environments

## When to avoid
- you need maintained software - it is deprecated in favor of OpenAI Retro
- you want modern TensorFlow or PyTorch versions
- you need production RL training infrastructure

## Facets
- artifact type: learning-resource
- maturity: abandoned
- function: reinforcement-learning, machine-learning, agent-framework, benchmarking
- domain: reinforcement-learning, machine-learning, artificial-intelligence, gaming-tools, tutorials
- platform: python
- tags: a3c, openai-gym, universe-environments, atari, starter-agent, deprecated, tensorflow, linux, macos, gpu

## Member repositories
- openai/universe-starter-agent (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:34.691977+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:46:36.780486+00:00, confidence not recorded.
  - readme: https://github.com/openai/universe-starter-agent (fetched 2026-08-28T04:03:34.691977+00:00, sha 949923a9d9ba)
- Data as of 2026-08-30T08:39:29.467469+00:00.