What is nemo-gym-release-publish?

A release procedure for publishing a NeMo Gym version, including a dry run, installation check, and release to PyPI and GitHub. PyPI is a package registry used to distribute Python software.

Use it to validate a release workflow, check the built wheel, publish a specified commit, and verify the PyPI and GitHub results.

It reduces the risk of publishing the wrong commit or a package that cannot be installed. It also requires approval before the permanent upload step.

How NVIDIA-NeMo/Gym configures Claude Code and Codex

About the project

NVIDIA-NeMo/Gym is a library and infrastructure for evaluating and improving models and agents inside environments, where each environment defines tasks, agent interaction, verification, and execution state. It is for teams running reproducible evaluations or training at scale across settings such as code execution, tool calling, and sandboxes, and the catalogue entries provide skills and instructions for working with it.

NVIDIA-NeMo/Gym · 1,320 stars · on GitHub · docs.nvidia.com

Reuse

Borrowing it

Nothing to install: this file belongs to NVIDIA-NeMo/Gym. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/NVIDIA-NeMo/Gym/main/.agents/skills/nemo-gym-release-publish/SKILL.md

Clone the repo

git clone --depth 1 https://github.com/NVIDIA-NeMo/Gym

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/nvidia-nemo/gym/nemo-gym-release-publish/github.svg)](https://agentmods.dev/skills/nvidia-nemo/gym/nemo-gym-release-publish)

Your own site

<a href="https://agentmods.dev/skills/nvidia-nemo/gym/nemo-gym-release-publish"><img src="https://agentmods.dev/badge/skills/nvidia-nemo/gym/nemo-gym-release-publish/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/nvidia-nemo/gym/nemo-gym-release-publish"><img src="https://agentmods.dev/badge/skills/nvidia-nemo/gym/nemo-gym-release-publish.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 108 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 2,117 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00108	$0.02117
Opus 5.5	$0.00043	$0.00847
Sonnet 5.5	$0.00022	$0.00423
Haiku 4.5	$0.00011	$0.00212

Measured 22d ago against content hash 73175c8aa537, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

nemo-gym-release-publish scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 22d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -s -o /dev/null -w '%{http_code}\n' https://pypi.org/pypi/nemo-gym/$VERSION/json

.agents/skills/nemo-gym-release-publish/SKILL.md · 189 lines

How it starts

The opening of the file, as written. The whole thing — 189 lines — stays where its author put it; the contents beside it link to each section on GitHub.

NeMo Gym Release Publish

Publish one exact commit to PyPI and GitHub. Always run a dry run first, check that the wheel installs, and get the user's explicit go-ahead before you publish. A PyPI upload cannot be undone or overwritten.

Workflow facts

Workflow: .github/workflows/release.yaml, which calls the shared _release_library.yml template from NVIDIA-NeMo/FW-CI-templates.
Trigger it from main. The release-ref SHA decides what gets released, not the branch you trigger from.
The template restricts the run to repository admins.
skip-test-wheel: true is set, so the workflow never installs the wheel. The install check below fills that gap.
After publishing, the workflow pushes a version bump to version-bump-branch. A run with release-ref set to a branch name races that bump, so release-ref must be a full 40-character SHA.

1. Resolve the inputs

Take each value from the user's request. Derive a default only when the user did not give one, and ask when VERSION is missing.

Variable	Workflow input	Default
`VERSION`	none	Required. The version being published, such as `<MAJOR>.<MINOR>.<PATCH>`
`REPO`	none	`NVIDIA-NeMo/Gym`
`BRANCH`	`version-bump-branch`	`r$VERSION`
`SHA`	`release-ref`	Tip of `BRANCH`
`PREVIOUS_TAG`	`gh-release-from-tag`	Highest release tag below `v$VERSION`
`CREATE_GH_RELEASE`	`create-gh-release`	`true`
`PYTHON`	none	`3.13`, for the install checks

Read state from GitHub, not from local refs. A local checkout may not have fetched the release branch, and origin may be a fork.

REPO=NVIDIA-NeMo/Gym
VERSION=<VERSION>
BRANCH=r$VERSION
CREATE_GH_RELEASE=true
PYTHON=3.13

SHA=$(gh api repos/$REPO/branches/$BRANCH -q .commit.sha)
PREVIOUS_TAG=$(gh release list -R $REPO --limit 50 --json tagName -q '.[].tagName' \
  | { cat; echo "v$VERSION"; } | sort -uV | grep -B1 -x "v$VERSION" | head -1)

If the user gave a SHA, check that it is a full 40-character SHA on BRANCH. The compare status must be identical or ahead:

Read the full file on GitHub · 189 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

22d ago First seen · 189 lines · 108 tokens per session scan A 73175c8aa537

Subscribe to this mod's changes

nemo-gym-release-publish is a skill published in the GitHub repository NVIDIA-NeMo/Gym (1,320 stars, last pushed today), licensed Apache-2.0. It adds 108 tokens to every session and 2,117 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-16.

CI and CD DevOps and cloud Git and workflow Releases agents gym reinforcement-learning reinforcement-learning-environments rl-environment rl-training

nemo-gym-release-publish next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
nemo-gym-release-publishNVIDIA-NeMo/Gym	Skill	1.3k	today	Apache-2.0	A	108	Claude Code, Codex
swarm-migrateyonatangross/orchestkit	Skill	299	yesterday	MIT	A	103	Claude Code
afk-releasegriffinwork40/agent-afk	Skill	62	yesterday	Apache-2.0	A	156	Claude Code
shipdarkroomengineering/cc-settings	Skill	52	yesterday	MIT	A	45	Claude Code
zb-release-pipelineAdamBien/airails	Skill	62	yesterday	MIT	A	148	Claude Code, Codex
land-and-deployFlorianBruniaux/claude-code-plugins	Skill	44	yesterday	MIT	A	24	Claude Code
aiox-devopsSynkraAI/aiox-core	Skill	3.1k	yesterday	none	A	51	Claude Code

Gym: Skill for Claude Code