horizonrobotics/embodiedgen/layout-creator
Claude Code and agent skill by horizonrobotics/embodiedgen
- community
Checking that the SKILL.md file is still in the repository.
Generate interactive 3D layouts from task descriptions with EmbodiedGen using layout-cli. Use this skill whenever users ask to build task-driven 3D scenes, batch-generate layouts from task files, tune layout generation retries/seeds, or produce simulator-ready layout outputs from background scene lists.
Install
npx vskill@latest install horizonrobotics/embodiedgen/layout-creatorSKILL.md
Layout Creator
Unified entry for EmbodiedGen interactive layout generation via layout-cli.
When To Use
Use this skill when users want to:
- Generate interactive 3D scenes from task descriptions.
- Batch-generate layouts from a task list file.
- Build simulator-ready layout outputs (
layout.json, renders) with optional robot insertion. - Tune generation quality and stability via retry and seed settings.
Routing Rule (Core)
Use layout-cli when the user input is task-level natural language descriptions (e.g., "put the pen in the mug") and the target output is an interactive layout scene, not standalone assets or standalone background scenes.
Pre-checks
- Run commands from the repository root.
- Confirm the active environment is
embodiedgen. - Confirm background scene list file exists and is readable (via
--bg_list). - If CLI commands are unavailable, run
pip install -e .to register entrypoints.
Standard Command Templates
1) Generate layouts from inline task descriptions
2) Batch generation from task list file (background run)
Common Parameters
--task_descs: task descriptions or a task-list text file path.--output_root: root output directory.--bg_list: background scene list file (scene retrieval pool).--insert_robot: include robot pose in layout generation/simulation output.--output_iscene: export composed scene mesh (Iscene.glb).--n_image_retry --n_asset_retry --n_pipe_retry: retry controls for text-to-3D subpipeline.--seed_img --seed_3d --seed_layout: reproducibility controls.--n_img_sample --text_guidance_scale --img_denoise_step: text-to-image / asset-generation controls.--keep_intermediate: keep intermediate files from generation substeps.
Output Conventions
Outputs are organized by task index:
<output_root>/task_0000/layout.json<output_root>/task_0000/scene_tree.jpg<output_root>/task_0000/background/<output_root>/task_0000/asset3d/- Optional:
<output_root>/task_0000/Iscene.glb(when--output_isceneis enabled)
Runtime Expectations
- Typical generation time is around 30 minutes per task (depends on retries/GPU/background matching).
- Batch jobs should use background execution (
nohup) with log redirection.
Failure Handling and Retry
- Missing background candidate: verify
--bg_listpath and referenced scene directories. - OOM or GPU pressure: reduce concurrency and lower retry/sample settings.
- Poor asset/layout quality: increase retry counts or refine task text.
- Missing outputs: verify output permissions and use absolute paths.
Security
- Security scanPassed, 100 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingembodied_gen/skills/layout-creator/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/horizonrobotics/embodiedgen/layout-creator)