An AI-powered vision skill that identifies and highlights specific delivery packages on crowded shelves using pickup codes.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install find-package
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 find-package 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 find-package?
The find-package skill is a sophisticated computer vision utility designed for Openclaw Skills users to navigate the chaos of delivery stations. By processing user-provided pickup codes and shelf photographs, the skill identifies the exact location of a package and provides visual confirmation with annotated bounding boxes. It is specifically optimized for Chinese delivery environments, such as Cainiao stations and Hive Box lockers.
This tool leverages advanced OCR and image processing to bridge the gap between digital delivery notifications and physical package retrieval. By automating the visual search process, Openclaw Skills allow users to save time and reduce frustration when dealing with hundreds of similar-looking parcels on high-density shelving units.
find-package Use Cases
- Searching for specific packages in a crowded delivery station or station locker.
- Extracting pickup codes automatically from SMS notifications or app screenshots.
- Highlighting package locations for users with visual impairments or difficulty reading small labels.
- Managing multiple package retrievals simultaneously by tracking found and missing items.
How find-package Works
- The AI extracts a pickup code from user text or a screenshot of a delivery app notification.
- The user provides one or more photos of the delivery shelves in the station.
- The skill uses vision capabilities to scan labels and match them against the confirmed pickup code.
- A specialized Python script generates an annotated version of the photo with a red bounding box around the target package.
- The skill sends the annotated image and location confirmation back to the user via the designated messaging channel.
find-package Setup
To use this within the Openclaw Skills framework, ensure you have python3 installed on your host system. The skill requires the annotation script located in your local directory.
# Ensure python3 is available
python3 --version
# The skill uses the following internal annotation command structure:
python3 {baseDir}/scripts/annotate.py --input /path/to/shelf_photo.jpg --output /tmp/result.jpg --box "x1,y1,x2,y2"
find-package Data Schema & Taxonomy
The skill organizes data dynamically during the session to track retrieval progress:
| Data Type | Description | Format |
|---|---|---|
| Pickup Code | Extracted reference number (Shelf-Row-ID) | String (e.g., 5-2-1234) |
| Input Media | Original shelf photos provided by the user | Image File (.jpg, .png) |
| Annotated Result | Processed image with highlighted bounding boxes | Image File (stored in /tmp/) |
| Status Tracking | Counter for found vs. missing packages | Integer/List |
find-package Advanced Features
- Automated OCR extraction from delivery app screenshots (Cainiao, Fengnest, etc.).
- Multi-package tracking for batch pickups with real-time status reporting.
- Dynamic image annotation using local Python scripts for privacy and speed.
- Support for varied pickup code formats including X-Y-ZZZZ and pure numeric strings.
- Integrated Telegram messaging support for mobile-first user interaction.
SKILL.md
Loading
Related Openclaw Skills
Workout Tracker AI Agent Skill
An intelligent fitness logging skill that parses natural language workout data into a local MySQL database for progress tracking and smart training plans.
HumanAI Convention
A participation skill for AI agents to validate semantic grounding and receive cryptographic participation receipts via the HumanAI Convention protocol.
Douyin Video Downloader & Analyzer
A comprehensive tool for extracting watermark-free Douyin videos, performing AI-powered transcriptions, and conducting deep content analysis via Openclaw Skills.
Lessie — People Search & Enrichment
Lessie is a comprehensive intelligence tool for discovering, qualifying, and enriching professional profiles and organizational data.
ScrapingBee Fetch
An advanced web extraction tool that renders JavaScript-heavy websites and returns clean Markdown optimized for LLM processing.
Free Subtitle Maker
An AI-powered skill that transcribes video audio and burns styled subtitles directly into the video file for social media and accessibility.
Featured*
