A precision skill for searching, validating, and pulling off-the-shelf catalog component STEP files directly into CAD workflows.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install step-parts
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install step-parts using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is step-parts?
The step-parts capability empowers AI agents to find, evaluate, and download common purchasable CAD components from the hosted step.parts catalog. This prevents agents from wasting compute on generating simplified placeholder geometry or scraping HTML when exact off-the-shelf actuators, servos, motors, electronics boards, connectors, and fasteners are readily available as canonical assets.
By serving as a bridge to the step.parts REST API, this addition to Openclaw Skills ensures precise matching of standards (such as ISO), aliases, and specific dimensions. It ensures that standard components are perfectly specified and checksum-verified before they enter the downstream assembly pipeline.
step-parts Use Cases
- Searching for precise hardware components like screws, bolts, nuts, and bearings based on specific standards or dimensions.
- Locating exact servo and actuator models using fuzzy tokens, brand aliases, or part families (e.g., Feetech, Waveshare).
- Fetching canonical .step files to substitute generic placeholders in a 3D assembly design.
- Automating the download and validation of hardware files within localized sandbox environments.
How step-parts Works
- Interpretation: The agent parses user requests or assembly specifications into descriptive queries, exact facets (category, family, standard), or dimensional attributes.
- Catalog Query: The skill sends a structured request to the canonical
https://api.step.parts/v1/partsendpoint, exploring dropped letters or vendor aliases if the initial lookup returns empty. - Resolution & Selection: Ambiguous options are evaluated based on their attributes (e.g., length, thread, bore size) to select the optimal component record.
- Download & Checksum Verification: The agent retrieves the component using its dedicated
stepUrland runs a SHA-256 verification via the bundled utility script to confirm asset integrity. - Visualization Handoff: Once saved locally, the file path is immediately forwarded to the CAD Viewer tool to update or render the workspace.
step-parts Setup
The skill provides a dedicated Python pipeline for headless environments. Ensure your runtime environment has appropriate outbound network access configured.
# Search for standard hardware using text queries
python scripts/download_step_part.py "M3 socket head 12" --download
# Download a specific catalog asset via its unique ID
python scripts/download_step_part.py --id iso4762_socket_head_cap_screw_m3x12 --download
# Query and limit search results for discovery inspection
python scripts/download_step_part.py "bearing 608zz" --limit 5
step-parts Data Schema & Taxonomy
The skill interacts with the remote schema and tracks local storage path metadata.
API Endpoints & Objects
| Endpoint | Target Resource / Purpose |
|---|---|
/v1/parts |
Main filtered search index returning full item lists and available asset facets. |
/v1/parts/{id} |
Fetches an individual enriched component entry with precise metadata and absolute asset paths. |
/v1/catalog/parts.index.json |
Compact catalog index minimized for swift client-side discovery. |
/v1/catalog/schema |
Field definition mappings and family-specific attribute taxonomies. |
Component Attributes Taxonomy
The attributes payload handles specific physical dimensions and specifications directly:
- Fasteners:
thread,lengthMm,material,standard.designation - Motion Components:
bore1Mm,profileSeries,slotSizeMm, voltage, torque rules.
step-parts Advanced Features
- CAD Viewer Handoff Protocol: Automatic background routing triggers (
$cad-viewer) to refresh interactive workspaces as soon as new files hit the disk. - LFS Environment Adaptability: Seamless fallback transitions between CDN asset delivery nodes and backend GitHub LFS media streams depending on workspace orientation.
- Bulk Page Pulls: Supports an
--allcompilation directive to capture entire query page returns cleanly in a single automated pass. - Fuzzy Alias Resolution Engine: Smart mitigation layers that intercept dropped naming syntax or manufacturer rebrands (e.g., mapping
STS3215to its full vendor tree).
SKILL.md
Loading
Related Openclaw Skills
DXF Generation and Validation
An AI agent skill to programmatically generate, update, and validate 2D DXF drawings and CAD projections from Python sources.
TKSeller
TKSeller is an interactive automation skill designed to integrate custom TikTok seller SaaS platforms directly with messaging clients.
novel-master
A publication-grade Chinese novel writing skill pack designed for long-form fiction using a distributed multi-agent architecture.
biaoshu-writer-corp
An AI-powered bidding assistant that automates document parsing, docx generation, and strict compliance risk analysis via an asynchronous API connection.
LegionSpace Version Checker
An Openclaw Skills utility that checks LegionSpace version availability across seven Chinese app stores and exports the latest results automatically.
Claw Markdown Preview
A local Markdown preview server enabling real-time document rendering, side-by-side editing, and rich text export to publishing platforms.
Featured*
