Skip to content
Back to skills

Robot Model Import

ASecurity

"Load, inspect, validate, and troubleshoot IKPy robot models from

  • 247 stars
  • 0 votes
  • 0 copies
  • 3 views
  • Added September 8, 2026
toolspythonapibackend

Works with

  • api

Security analysis

A100/100

Pro scans all 6 files and shows the line behind each finding

Scanned September 8, 2026

npx -y skills add VectorSpaceLab/AREX-Skill --skill robot-model-import --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Robot Model Import?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Robot Model Import
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/vectorspacelab-robot-model-import/badge)](https://www.skillsdirectory.com/skills/vectorspacelab-robot-model-import)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: robot-model-import
description: "Load, inspect, validate, and troubleshoot IKPy robot models from
  URDF, MJCF, or IKPy JSON metadata."
disable-model-invocation: true
metadata:
  disco-role: operating
license: Apache 2.0
---

# Robot model import

Use this sub-skill when a robot description must become an IKPy `Chain`, when
an imported chain has the wrong path or joint count, or when a URDF/MJCF parser
error needs diagnosis. Start with the smallest model inspection that answers
which links, joints, bodies, and path names are present.

- For loader signatures, path semantics, element-to-`URDFLink` mapping, and
  orientation conventions, read [api-reference.md](references/api-reference.md).
- For self-contained URDF/MJCF examples and smoke fixtures, read
  [workflows.md](references/workflows.md). The bundled
  `scripts/make_tiny_models.py` creates temporary examples without repository
  assets.
- For actionable parser and data failures, read
  [troubleshooting.md](references/troubleshooting.md).
- `python scripts/inspect_model.py --help` provides a no-write model inspection
  command. It reports source names and chain shape, and can optionally run FK.

The central distinction is structural: URDF has root-level `link` and `joint`
records traversed as an alternating path, while MJCF nests `body` records and
follows child bodies. IKPy represents each URDF/MJCF joint (or a fixed
body/explicit tip offset) as a `URDFLink`; the source URDF link records are not
chain links. Keep explicit `False` values in an `active_links_mask` for the
origin and fixed tip entries; do not rely on the constructor's last-entry
warning to repair a NumPy boolean mask. Pass a full joint vector, including
inactive entries.

Use `Chain.from_urdf_file` or `Chain.from_json_file` for URDF-backed models and
`Chain.from_mjcf_file` for MJCF. Use `ikpy.urdf.utils.get_urdf_tree` only for
optional structural inspection, and keep rendering concerns in
`visualization-geometry`. Route FK/IK method selection to `chain-kinematics`
and JAX execution to `jax-backend`.

Files in this skill

  • SKILL.md2 KB
  • references/api-reference.md12.2 KB
  • references/troubleshooting.md11.7 KB
  • references/workflows.md7.9 KB
  • scripts/inspect_model.py6.2 KB
  • scripts/make_tiny_models.py3.5 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.