meta-pytorchOfficial·15 Agent Skills Included

OpenEnv

Build, deploy, and train AI agents in isolated RL environments

Creates isolated, containerized execution environments where AI agents can act, learn, and be trained with reinforcement learning. Provides simple Gymnasium-style APIs (reset, step, state) plus a CLI to scaffold, validate, and deploy environments to Hugging Face Spaces. Removes the pain of wiring servers, Docker images, and WebSocket protocols by hand, so teams ship training-ready environments faster.

2,537|432

↗

npx skills add meta-pytorch/OpenEnv --all -g -y

Available:

Tells the AI agent how to work inside this repository: which skills and subagents to trigger for TDD workflows, code review, environment generation, and deployment, plus the exact build, test, and lint commands to run.

All Skills in This Repository (15)

Pure Emerald Level Indicators

📦 In Repo

meta-pytorch

hf-cli

Manages Hugging Face Hub repositories, models, datasets, and Spaces via the hf command-line interface.

openenv-cli

Scaffold, validate, build, and push OpenEnv agentic execution environments via the openenv CLI.

watch-pr

Monitor GitHub PR CI checks and Greptile reviews, auto-fixing failures until green.

update-docs

Detects stale documentation references after API changes and propagates fixes across the repository.

release

Deploys OpenEnv environments to Hugging Face Spaces and syncs canonical collections.

hf-space-recovery

Diagnose and recover failing Hugging Face Space deployments for OpenEnv environments.

sprint

Coordinates parallel GitHub issue resolution using agent teams, worktrees, TDD, and stacked PRs.

generate-openenv-env

Generate OpenEnv reinforcement learning environments from a use-case description.

work-on-issue

Parse GitHub issues into goals, acceptance criteria, and constraints.

alignment-review

Assess code changes for alignment with OpenEnv principles and RFCs.

rfc-check

Determine if proposed changes require an RFC by applying criteria and checking conflicts.

write-tests

Generate failing test scaffolds from todos or requirements for TDD workflows.

Official

Intermediate

Frequently Asked Questions

FAQPage Schema

How to install OpenEnv?▼

Run `npx skills add meta-pytorch/OpenEnv --all -g -y` in your terminal to install all skills in this suite globally.

What is OpenEnv used for?▼

OpenEnv is a framework for creating and deploying isolated execution environments where AI agents take actions and receive rewards, mainly for reinforcement learning training loops.

How do I create a new RL environment with OpenEnv?▼

Use the openenv CLI to scaffold a typed environment with models, client, server, and Dockerfile, then validate and deploy it to Hugging Face Spaces with built-in commands.

Does OpenEnv work with Claude Code and other AI agents?▼

Yes. The repository ships agent skills and orchestration files that let Claude Code and compatible agents run TDD workflows, reviews, and deployments automatically.

Can I deploy OpenEnv environments without Docker knowledge?▼

Yes. The CLI and deployment skills handle Docker images and Hugging Face Space publishing for you, so you only define the environment logic.