marin: Skill for Codex

.agents/skills/launch-rl/SKILL.md

launch-rl is a skill for Codex from marin-community/marin. It costs 58 tokens per session (736 once invoked), scanned A, original, Apache-2.0.

A workflow for defining, checking, submitting, and restarting reinforcement-learning experiments on the Marin SkyRL system. Reinforcement learning trains a model through feedback from an environment or evaluation process.

In plain words
What is it for?
Use it to create SkyRL experiment launch files, configure roles and hardware placement, run smoke tests, validate generated settings, and launch or relaunch experiments on Iris.
Why use it?
It keeps experiment settings, model and data dependencies, and cluster placement in the system's expected configuration instead of duplicating them across launch commands. It also requires validation before a real run.

Skill for Codex

Written for Codex: agents/openai.yaml present. Also seen: installed under .agents/ (shared by several agents).

This is marin-community/marin's own configuration. It tells Codex how to work on marin itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything marin configures →

About the project

Marin is an open-source research program, software platform, and community for developing foundation models such as large language models. Researchers use it for data preparation, tokenization, pretraining, posttraining, evaluation, and related experiments, including work on audio-text, DNA, and protein models. The catalogue entries are add-ons that support workflows around Marin.

marin-community/marin · 4,074 stars · on GitHub · marin.community

Reuse

Borrowing it

Nothing to install: this file belongs to marin-community/marin. 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/marin-community/marin/main/.agents/skills/launch-rl/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/marin-community/marin

Made for: 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.

agentmods badge for launch-rl

README.md
[![agentmods](https://agentmods.dev/badge/skills/marin-community/marin/launch-rl/github.svg)](https://agentmods.dev/skills/marin-community/marin/launch-rl)
Your own site
<a href="https://agentmods.dev/skills/marin-community/marin/launch-rl"><img src="https://agentmods.dev/badge/skills/marin-community/marin/launch-rl/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.

agentmods 80×15 button for launch-rl

Your own site · 80×15
<a href="https://agentmods.dev/skills/marin-community/marin/launch-rl"><img src="https://agentmods.dev/badge/skills/marin-community/marin/launch-rl.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 58 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 736 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00058 $0.00736
Opus 5.5 $0.00023 $0.00294
Sonnet 5.5 $0.00012 $0.00147
Haiku 4.5 $0.00006 $0.00074

Measured 15d ago against content hash b3fb79765ba1, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

launch-rl scanned grade A with 0 findings 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 15d 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.agents/skills/launch-rl/SKILL.md · 68 lines

How it starts

The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Launch RL

Use the experiment's artifact graph as the launch interface. For the configuration model, role arithmetic, and ownership boundaries, read RL launching.

Define the experiment

  • Define a Click main that constructs and returns the requested terminal ArtifactStep handles.
  • Decorate it with @rl_build_options. Do not add another submission path to an experiment.
  • Put artifact identity and a fully explicit SkyRLRolePlan in SkyRLSpec. Put cluster placement in IrisSkyRLExecution.
  • Render role geometry from the plan. Do not repeat DP, TP, PP, EP, engine count, placement, or batch sizes as independent literals or Hydra overrides.
  • Express models and data as artifact dependencies. Pin tokenizer and source revisions; do not depend on an ambient checkpoint path or whatever happens to be on a worker.
  • Keep the generated SkyRLLaunchConfig YAML as the only Marin-to-MarinSkyRL boundary. Extend that config instead of adding request envelopes or internal per-field command-line flags.

Validate before launch

Run the main without --run first, using an immutable calendar version for a real experiment:

uv run python -m experiments.<module> --version YYYY.MM.DD <selection-options>

Treat a preflight failure as a recipe error. Fix the source config or role plan instead of weakening the validator. Check the Marin revision, physical allocation, role geometry, entrypoint, batch sizes, artifact identities, output paths, and terminal stage.

Use a fresh RL artifact version after runtime changes. The temporary checkpoint root follows the artifact name and version, so reusing a version can make resume_mode=latest load private Torch or distributed state written by the old runtime. Reuse an RL version across runtime changes only when checkpoint compatibility has been established explicitly or resume points at a fresh checkpoint root.

Before a large launch, run a smoke through the same strategy, artifact handoffs, and role geometry. Confirm the runtime contains the required backend fixes and that host resources, credentials, context budget, checkpoint/export policy, and W&B identity are explicit.

Read the full file on GitHub · 68 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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.

  1. 15d ago Changed b3fb79765ba1
  2. 17d ago Changed 7e48546d4b16
  3. 1mo ago First seen · 68 lines · 58 tokens per session scan A ff5681febf0a

Subscribe to this mod's changes

launch-rl is a skill published in the GitHub repository marin-community/marin (4,074 stars, last pushed today), licensed Apache-2.0. It adds 58 tokens to every session and 736 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-24.

Related

launch-rl compared with skills from other repositories

launch-rl next to 6 skills from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
launch-rlmarin-community/marinSkill4.1ktodayApache-2.0A58Codex
instrument-data-to-allotropeanthropics/knowledge-work-pluginsSkill35k2d agoApache-2.0A123Claude Code
exploratory-data-analysisK-Dense-AI/scientific-agent-skillsSkill56k4d agoMITA83Claude Code
matlabK-Dense-AI/scientific-agent-skillsSkill56k4d agoMITA44Claude Code
phylogeneticsK-Dense-AI/scientific-agent-skillsSkill56k4d agoMITA71Claude Code, Codex
research-engineerdavila7/claude-code-templatesSkill33ktodayMITA43Claude Code, Codex
mapping-to-snomedmaziyarpanahi/openmedSkill5.8kyesterdayApache-2.0A205Claude Code