Add Robot logo

Add Robot

OrganizationPopular
AgibotTech
add-robot

Bring a custom robot into the Genie Sim RT Engine — author / fix a xacro / URDF in `genie_sim_robot_model`, prep meshes with the offline tools (`normalize_obj_names.py`, `diagnose_urdf.py`, `recompute_inertia.py`, `fix_dae_units.py`, `copy_dae_material.py`), stage assets into the AS3 layout (`robot.usda` + `payloads/Physics/{physics,physx,mujoco}.usda`) that the engine consumes, and wire the result into a `scene_*.yaml`. Trigger: When the user asks to "add a new robot", "import a robot", "support <vendor> in geniesim", names a URDF / xacro not currently in `genie_sim_robot_model/urdf/`, or wants to convert a third-party mesh pack into AS3-ready USD.

Overview

PublisherAgibotTech
Repositorygenie_sim
Skill nameadd-robot
Stars
1.4K
Forks
119
Bundled files
Instructions only
LicenseMPL-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by AgibotTech on GitHub. Read the source before you install it.

Installation

Install the Add Robot AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/AgibotTech/genie_sim.git /tmp/genie_sim
mkdir -p .claude/skills
cp -r /tmp/genie_sim/source/geniesim_ros/skills/add-robot .claude/skills/add-robot
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Add Robot in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Add Robot on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Add Robot is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

When to Use

  • User wants to run an off-tree robot on the RT Engine.
  • User has a URDF / xacro that fails to import (missing inertia, unit mismatch, DAE materials lost, OBJ names colliding) and needs the offline mesh-prep tools.
  • User wants to add the new robot to a scene_*.yaml and launch it via the launch-scene skill.

Do not use for:

  • Tweaking an existing Genie G2 variant → just edit the relevant URDF / xacro and rebuild via build-workspace.
  • Authoring a MoveIt config for the new robot → that's a separate, larger job (port SRDF + coupled_constraints.yaml).
  • Adding a benchmark task that uses the new robot → run-benchmark in geniesim_benchmark/skills/.

Critical Patterns

  1. Tier 1 only validates Genie G2. Reference robots (Franka, UR5, Aloha, ARX, Agilex) have correct URDFs and import cleanly, but scene yamls and physics tuning may be stale. Treat them as starting points, not certified configs.
  2. AS3 layout is mandatory. The engine expects the per-robot directory robot.usda + payloads/Physics/{physics,physx,mujoco}.usda regardless of which backend is active — switching physics never rewrites the asset.
  3. Fix the URDF before importing. diagnose_urdf.py catches missing inertia, zero-mass links, collision/visual mismatches, and bad OBJ name collisions — every one of these breaks the URDF→USD importer downstream.
  4. Mesh names must be unique. USD's flat namespace will silently merge two meshes that share an OBJ name. normalize_obj_names.py rewrites names to be prefix-unique.
  5. Mimic joints handled at runtime. Don't model gripper mimics inside the URDF — the engine bridges them via the controller layer (see 86308f040 in git history). URDF stays plain.
  6. Robots live in genie_sim_robot_model, not in geniesim_assets. Meshes can move to geniesim_assets once stabilised, but URDFs + xacros stay in the ROS package.

Workflow

Step 1 — Stage the URDF / xacro

Drop the file under source/geniesim_ros/src/ros_ws/src/genie_sim_robot_model/urdf/ next to the reference robots. Name follows the convention <vendor>_<model>[_<gripper>].urdf.

bash
ls source/geniesim_ros/src/ros_ws/src/genie_sim_robot_model/urdf/
# e.g. franka_fr3.urdf, ur5_robotiq_140.urdf, your_robot.urdf

Step 2 — Diagnose

bash
cd source/geniesim_ros/src/ros_ws/src/genie_sim_robot_model
python scripts/diagnose_urdf.py urdf/<your_robot>.urdf

Look for: missing <inertial>, zero mass, visual/collision link mismatches, mesh path resolution failures.

Step 3 — Prep meshes

For each issue the diagnose tool flagged:

bash
# OBJ name collisions:
python scripts/normalize_obj_names.py path/to/meshes/

# DAE units off (mm vs m):
python scripts/fix_dae_units.py path/to/meshes/<file>.dae

# DAE materials missing after a re-export:
python scripts/copy_dae_material.py source.dae target.dae

# Missing or wrong inertials (uniform-density approximation):
python scripts/recompute_inertia.py urdf/<your_robot>.urdf

Step 4 — Confirm the URDF imports

Re-run diagnose_urdf.py until it's clean, then build the workspace (build-workspace skill) and let the engine's URDF→USD importer stage the AS3 layout on first run. The importer writes:

assets/scenes/<scene>/
├── manifest.json
├── scene.usda
└── robot/
    ├── robot.usda
    └── payloads/
        └── Physics/
            ├── physics.usda
            ├── physx.usda
            └── mujoco.usda

manifest.json presence = cache hit. To force a regenerate:

bash
rm -rf assets/scenes/<scene>/
# or pass:
ros2 launch genie_sim_bringup app.launch.py … always_regenerate_robot_usd:=true

Step 5 — Wire into a scene yaml

Create genie_sim_bringup/config/scene_flat_<your_robot>.yaml modelled on one of the reference scenes (e.g. scene_flat_fr3.yaml). Required keys:

yaml
robot:
  urdf: <your_robot>.urdf            # filename, resolved against genie_sim_robot_model/urdf/
  init_base_pose:                    # non-physics teleport at spawn
    x: 0.0
    y: 0.0
    z: 0.0
    theta: 0.0
  init_joint_pos:                    # optional, per-joint dict
    joint_name: <value>

viewer_camera:                       # Newton GL viewer default cam
  pos: [1.6, -1.6, 1.2]
  lookat: [0.0, 0.0, 0.8]

scene:
  base_path: <relative_to_geniesim_assets>
  scene_usda: <stage>.usda           # or omit for an empty scene

Step 6 — Launch via launch-scene

bash
ros2 launch genie_sim_bringup app.launch.py \
  scene:=scene_flat_<your_robot> \
  launcher_config:=launcher_ovrtx_isaac_physx \
  headless:=false   # local workstation with a screen; flip to true on a remote/headless host

If anything goes wrong, fall back to the experimental launcher_newton_mjwarp to see whether the issue is PhysX-specific or applies to all backends.

Commands (copy-paste summary for the user)

bash
# Inside the container:
cd source/geniesim_ros/src/ros_ws/src/genie_sim_robot_model

# 1. Diagnose
python scripts/diagnose_urdf.py urdf/<your_robot>.urdf

# 2. Mesh prep (only the ones the diagnose tool flagged)
python scripts/normalize_obj_names.py path/to/meshes/
python scripts/fix_dae_units.py path/to/meshes/<file>.dae
python scripts/copy_dae_material.py src.dae dst.dae
python scripts/recompute_inertia.py urdf/<your_robot>.urdf

# 3. Build, then let the engine stage AS3 on first run
cd /workspace
geniesim ros build dev && source devel/setup.bash

# 4. Author scene_flat_<your_robot>.yaml, then:
ros2 launch genie_sim_bringup app.launch.py \
  scene:=scene_flat_<your_robot> \
  launcher_config:=launcher_ovrtx_isaac_physx \
  headless:=false   # local workstation with a screen; flip to true on a remote/headless host

Notes

  • The AS3 layout is the same regardless of backend, so a robot that works under Isaac PhysX also works under Newton-standalone — only the physics tuning (gains, contact compliance) may need per-backend trimming.
  • recompute_inertia.py assumes uniform density and is only a starting point. For Tier 1 quality, source inertials from the vendor's CAD.
  • The URDF→USD importer is cache-gated by manifest.json. Editing init_joint_pos does not bust the cache (it's forwarded as a JSON-encoded launch param, not a manifest field) — only changes that affect topology (URDF, xacro args, mimic, fixed-base selection) require a regenerate.
  • For non-G2 robots, MoveIt is not packaged. The engine runs fine without it; you just don't get planning or RViz interactive markers out of the box.

Resources

Frequently asked questions

What does the Add Robot AI skill do?

Bring a custom robot into the Genie Sim RT Engine — author / fix a xacro / URDF in `genie_sim_robot_model`, prep meshes with the offline tools (`normalize_obj_names.py`, `diagnose_urdf.py`, `recompute_inertia.py`, `fix_dae_units.py`, `copy_dae_material.py`), stage assets into the AS3 layout (`robot.usda` + `payloads/Physics/{physics,physx,mujoco}.usda`) that the engine consumes, and wire the result into a `scene_*.yaml`. Trigger: When the user asks to "add a new robot", "import a robot", "support <vendor> in geniesim", names a URDF / xacro not currently in `genie_sim_robot_model/urdf/`, or...

Why use Add Robot on TypingMind?

Because you install it once and use it with any model. Add Robot is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Add Robot in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/AgibotTech/genie_sim/tree/main/source/geniesim_ros/skills/add-robot. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Add Robot?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Add Robot?

As many as you like. As long as a model supports skills, you can use Add Robot with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Add Robot AI skill free?

Yes. It is published on GitHub by AgibotTech under the MPL-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇