installation

There are four ways to install prime-rl, and which one you want depends on whether you are just running it, running the shrunk-down lower-precision versions of big models, or working on the code itself.

WHEN TO USE

SCOPE

USE FOR

Install the core prime-rl dependencies to get the project running.
Install every optional extra at once instead of adding them one at a time.
Install the extra package required to run FP8 models like GLM-5-FP8.
Install development tools such as the test runner and linter for contributing to the project.

This is a saved copy of the skill's SKILL.md instructions. It may differ from the upstream file or the version you install.

In this document (6)

Basic install

uv sync

This installs all core dependencies defined in pyproject.toml.

All extras at once

The recommended way to install for most users:

uv sync --all-extras

This installs all optional extras (flash-attn, flash-attn-cute, etc.) in one go.

FP8 inference with deep-gemm

For certain models like GLM-5-FP8, you need deep-gemm. Install it via the fp8-inference dependency group:

uv sync --group fp8-inference

This installs the pre-built deep-gemm wheel. No CUDA build step is needed.

Dev dependencies

uv sync --group dev

Installs pytest, ruff, pre-commit, and other development tools.

Key files

pyproject.toml — all dependencies, extras, and dependency groups
uv.lock — pinned lockfile (update with uv sync --all-extras)

Local tool required

1 FOR THE AGENT

01
Install uv to run the setup commands
All the install commands here (uv sync, uv sync --all-extras, uv sync --group ...) are run through uv, the Python package/project manager, so uv must be installed first.

Not standalone. This skill ships with huggingface/prime-rl and only works together with that tool — install the tool first, then add this skill.

Run this outside your AI tool, using Node.js 22.20.0 or newer. Skills CLI 1.7.2 skips skill selection; prompts cancel installation without an interactive terminal. If choosing Project when asked about installation scope, open your project's folder before running this.

Install checklist

Check selected tools if a picker appears; Claude Code may be unchecked under Additional agents. ↑/↓ moves, Space toggles selections, Enter confirms.
Without a tool picker, check Installing to: for your intended tool. Stop if it names a different tool.
If asked, choose Symlink (Recommended) for shared files or Copy to all agents for separate copies; review the summary before confirming installation.
Install the find-skills skill? defaults to Yes, adding another skill for all projects. Choose No if you don't want that extra installation.

Skills CLI 1.7.2 sends install data to skills.sh by default, including the source, skill names, target tools and file details; GitHub sources are reported only when GitHub confirms the repository is public, while other remote sources can also be reported. Confirmed-public GitHub sources also trigger a security-audit request. Before installing, set the environment variable DO_NOT_TRACK=1 or DISABLE_TELEMETRY=1 to disable both.

npx skills add huggingface/prime-rl --skill "installation" --full-depth

After installing for Claude Code, open it in the installation project if you chose Project; Global makes the skill available across projects on this computer. Once Claude Code loads the skill, its description helps Claude decide when to use it.

If installation created a top-level skills folder that did not exist when your Claude Code session started, run /reload-skills in that session. Run it again after later changes in that folder.

The repository is licensed under Apache-2.0 — use, modification, and redistribution are allowed under that license’s terms.

View the repository’s license file on GitHub →

installation

Installation

Basic install

All extras at once

FP8 inference with deep-gemm

Dev dependencies

Key files

Local tool required

More skills

inference-server

toml-config

MCP keyword matches

shadcn MCP Server

agent-device

shadcn/ui MCP

MCP Installer