What is open-dreamer?

A setup and operating guide for Open Dreamer, an open JAX/Flax implementation of the Dreamer 4 world-model training pipeline. The pipeline includes a video tokenizer, an action-conditioned model of changing scenes, generated rollouts, and FVD scoring.

Use it to package and run Open Dreamer through NPA’s BYOF workflow, test its tokenizer and dynamics on at least two GPUs, or onboard similar world-model solutions.

It records the project details and run requirements needed to onboard or debug this multi-GPU workload. It also provides a reference for testing a world-model workflow with synthetic data instead of the full dataset.

open-dreamer — Skill for Claude Code and Codex by nebius

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent

npx skills add nebius/nebius-physical-ai --skill open-dreamer

Clone the repo

git clone --depth 1 https://github.com/nebius/nebius-physical-ai

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/nebius/nebius-physical-ai/open-dreamer/github.svg)](https://agentmods.dev/skills/nebius/nebius-physical-ai/open-dreamer)

Your own site

<a href="https://agentmods.dev/skills/nebius/nebius-physical-ai/open-dreamer"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/open-dreamer/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/nebius/nebius-physical-ai/open-dreamer"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/open-dreamer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 64 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 3,043 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. Third-party audits

NVIDIA SkillSpector warn 7 Sept 2026

SkillSpector: 2 findings, up to medium

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

medium Data Exfiltration · line 98
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
medium Tool Misuse · line 183
Tool defaults are unsafe or overly permissive (e.g. disabled TLS verification, no authentication, world-writable permissions). Unsafe defaults widen the attack surface.

Fix: Override unsafe defaults with secure settings (verify=True, auth required, restrictive permissions). Review and harden all tool configurations.

How audits are shown

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	$0.00064	$0.03043
Opus 5	$0.00032	$0.01522
Sonnet 5	$0.00013	$0.00609
Haiku 4.5	$0.00006	$0.00304

Measured 3d ago against content hash ddc226e4da57, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

open-dreamer 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 3d 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 -sk -u "$AGENT_USER:$AGENT_PASSWORD" -X POST https://<agent-ip>/api/sim-viz/load-artifact \

skills/tools/open-dreamer/SKILL.md · 209 lines

How it starts

The opening of the file, as written. The whole thing — 209 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Open Dreamer (Dreamer 4 world model)

Open Dreamer (next-state/open-dreamer) is an open JAX/Flax NNX implementation of the Dreamer 4 world-model training pipeline: a causal video tokenizer, an action-conditioned latent dynamics model, rollout generation, and FVD scoring. It is onboarded into NPA as a multi-GPU BYOF registry candidate and is the reference example for the >=2 GPU BYOF path.

Load skills/workflows/byof-onboard/SKILL.md for packaging/run mechanics and skills/workflows/oss-solution-registry-onboard/SKILL.md for the registry admission contract. For onboarding a new world model, load skills/workflows/onboard-world-model/SKILL.md — the generic playbook for which this skill is the worked reference example. This skill records the solution-specific facts.

When To Use

Build/push and run the Open Dreamer registry image via BYOF solution-smoke.
Author or debug a multi-GPU (2-GPU minimum) world-model BYOF workflow.
Understand how to exercise the Dreamer 4 tokenizer/dynamics on synthetic data without staging the full Minecraft/VPT dataset.

Upstream Facts

Repo: https://github.com/next-state/open-dreamer · pinned ref 2b10640.
Stack: Python 3.11, uv, CUDA-12 jax[cuda12] + Flax NNX, Hydra/OmegaConf configs, Grain/ArrayRecord data. Install via uv sync.
Documented entrypoints (scripts/): train_tokenizer.py, tokenize_minecraft_dataset.py, train_dynamics.py, eval_fvd.py.
Multi-GPU: dreamer.parallel.build_parallel(strategy) builds a JAX device mesh over jax.devices() for data / fsdp / tp / sp. train_tokenizer forces data parallelism; train_dynamics uses parallel_strategy.
Datasets: raw Minecraft/VPT MP4 ArrayRecords (minecraft_vpt), a self- contained procgen coinrun path, and pre-tokenized latent ArrayRecords.

NPA Mapping

Item	Path
Workflow spec	`workflows/testing/byof-open-dreamer.yaml`
2-GPU resource profile	`npa/src/npa/workflows/byof/profiles/byof-solution-smoke-rtxpro-2gpu.yaml`
BYOF toolRef	`workbench.byof.repo` (`--workload solution-smoke`)
Catalog entry	`docs/workbench/oss-solution-catalog.md`

Read the full file on GitHub · 209 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.

3d ago Changed ddc226e4da57
11d ago First seen · 209 lines · 64 tokens per session scan A f670b9b49230

Subscribe to this mod's changes

open-dreamer is a skill published in the GitHub repository nebius/nebius-physical-ai (28 stars, last pushed today), licensed Apache-2.0. It adds 64 tokens to every session and 3,043 once invoked, about $0.0003 per session on Opus 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 2026-08-30.

Data and AI

agent-platform-rag-engine-management

Manage and query Agent Platform RAG Engine Corpora and retrieve grounded contexts using the Google GenAI SDK. Use when listing RAG corpora or files, inspecting a corpus, retrieving contexts, or generating content grounded in a RAG corpus. Do not use for standard database queries (use SQL/Spanner skills), Google…

google/skills · 85 tokens

agent-platform-model-registry

Agent Platform Model Registry Management. Use when you need to upload, list, describe, update, or delete machine learning models (and their versions) in the Agent Platform Model Registry. Don't use for model training, model deployment to endpoints, or managing non-Agent Platform models.

google/skills · 60 tokens

foundry-config-setup

Resolve missing setup caused by a hardcoded Foundry project endpoint or model in a sample. Use when a sample fails because it uses a placeholder/hardcoded projectendpoint (for example "https://your-project.services.ai.azure.com") or a hardcoded model instead of reading them from the environment.

microsoft/agent-framework · 65 tokens

google-cloud-solution-agentic-analytics-spark-knowledge-catalog

Discovers requirements and generates guidance to design and deploy a governed, secure agentic-analytics solution for data that's distributed across Google Cloud, other cloud providers, or on-premises. Data that's outside Google Cloud (such as data from Databricks, Snowflake, Salesforce, SAP, or Oracle systems) is…

google/skills · 138 tokens

training-check

Interactively monitor training metrics from the current Codex session, periodically checking WandB or fallback logs for NaN, divergence, plateaus, and broken runs.

wanshuiyin/Auto-claude-code-research-in-sleep · 35 tokens

nemo-automodel-launcher-config

Configure NeMo AutoModel job launches for interactive runs, Slurm clusters, and SkyPilot cloud execution.

NVIDIA/skills · 30 tokens

open-dreamer