A robust command-line toolkit for the comprehensive management, transformation, and validation of YOLO object detection datasets.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install yoloutils
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 yoloutils 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 YOLOUTils Agent Skill?
YOLOUTils is a developer-centric suite of Openclaw Skills designed to automate the heavy lifting of computer vision dataset preparation. It acts as a bridge between raw images and training-ready data by providing atomic commands for labeling, merging, and resizing. By operating directly on the source logic of yoloutils.py, this skill ensures that AI agents can perform high-fidelity data manipulations with technical precision.
The tool is particularly valuable for teams working with the YOLO architecture who need to maintain strict label consistency across distributed datasets. These Openclaw Skills empower developers to perform complex operations like model-assisted cropping and automated training-validation splits through simple, actionable commands.
YOLOUTils Agent Skill Use Cases
- Synchronizing and merging labels from disparate annotation sources into a unified project.
- Automated cleaning of datasets by removing specific label indices or empty annotation files.
- Batch processing image assets for specific input dimensions while preserving aspect ratios.
- Generating training-ready directory structures and data.yaml files from LabelImg-style exports.
- Running bulk inference tests on new datasets to generate confidence reports and CSV exports.
How YOLOUTils Agent Skill Works
- The AI agent analyzes the user's intent to select the appropriate subcommand from the ten available modules.
- A pre-execution check is performed to verify essential dependencies like opencv-python, ultralytics, and pyyaml.
- The skill maps the provided source and target directories, ensuring read/write permissions and the presence of classes.txt for label mapping.
- Processing logic is applied (e.g., recursive scanning, UUID renaming, or model-based cropping) with real-time feedback.
- Post-execution validation confirms the integrity of the output, such as the generation of the data.yaml file or statistical summaries.
YOLOUTils Agent Skill Setup
To utilize these Openclaw Skills, you must have the required Python environment configured. Install the core dependencies using the following command:
pip install opencv-python pyyaml pillow texttable tqdm ultralytics
You can then execute the skill via the entry point script:
# General usage
yoloutils <subcommand> --source <input_path> --target <output_path>
# Or via Python directly
python src/netkiller/yoloutils.py <subcommand> [arguments]
YOLOUTils Agent Skill Data Schema & Taxonomy
YOLOUTils follows standard YOLO data conventions while offering flexibility in output mapping. Data organization includes:
| Component | Organization Method | Metadata Generated |
|---|---|---|
| Dataset Splits | train/val/test subdirectories | data.yaml configuration |
| Labels | .txt files (YOLO format) | classes.txt mapping |
| Inference | Recursive target mapping | result.csv / visualization |
| Cropping | Class-based subfolders | Confidence scores |
YOLOUTils Agent Skill Advanced Features
- Model-driven cropping to automatically extract specific objects for classification training.
- UUID-based file renaming to ensure unique identifiers across large-scale merged datasets.
- Recursive directory traversal with smart filtering for system-specific files and metadata.
- Adjustable validation set sampling for precise control over training cycles.
- Batch search and replace functionality for re-indexing labels without manual file editing.
SKILL.md
Loading
Related Openclaw Skills
X Platform Skill
A specialized skill for AI agents to navigate and interact with X using route-aware workflows and robust verification rules.
Molty Royale Agent Skill
A specialized skill for AI agents to automate onboarding, gameplay, and reward management in the Molty Royale gaming ecosystem.
Publish Guard
Publish Guard is a specialized pre-release audit tool designed to scrub documentation and metadata for Openclaw Skills to ensure professional, leak-free public releases.
WeShop AI Ahegao CLI Skill
A command-line tool that uses AI to apply exaggerated anime-style facial expressions to portrait photos while preserving original details.
BaZi Fortune Analysis Master
A sophisticated AI agent specializing in traditional Chinese BaZi metaphysics to provide deep insights into life patterns, career, and relationships.
Million Dollar Landing Page
Transform functional SaaS landing pages into premium, high-converting digital experiences using the Million Dollar Landing Page framework for Openclaw Skills.
Featured*
