Skip to content

SKILL.md

Latest commit

 

History

History
105 lines (74 loc) · 3.15 KB

SKILL.md

File metadata and controls

105 lines (74 loc) · 3.15 KB
Raw
name photo-year-collage
description Creates a year-labeled photo collage from a folder of annual photos. Designed for China visa applications requiring facial-change documentation (历年生活照片), but works for any year-based photo timeline. Use when generating a chronological photo grid or visa life-photo proof (面相变化).
user-invocable true

Photo Year Collage

Creates a chronological photo collage grid with year labels from a folder of photos named by year. Handles HEIC, JPEG, and PNG formats automatically, including HEIC conversion on macOS.

When to Use

  • User needs to generate a collage of annual/yearly photos
  • China visa application requires proof of facial changes over time (面相变化)
  • User wants to rotate specific year photos before generating the collage
  • User provides a folder path or is already in a photo folder

Instructions

Step 1: Identify the Photo Folder

If the user hasn't specified a folder, use the current working directory. List the files to confirm they contain year-named photos (filenames starting with a 4-digit year).

Step 2: Handle Rotation Requests (Optional)

If the user wants to rotate specific year photos before generating the collage:

# Rotate clockwise 90° using macOS sips (handles HEIC, JPEG, PNG)
sips -r 90 "<filename>"

# Rotate counter-clockwise 90°
sips -r -90 "<filename>"

# Rotate 180°
sips -r 180 "<filename>"

Apply to all files matching the requested years, then confirm before proceeding to collage generation.

Step 3: Run the Collage Script

Use the bundled script:

SKILL_DIR="$(dirname "$(realpath ~/.claude/skills/photo-year-collage/SKILL.md)")"
python3 "$SKILL_DIR/scripts/create_collage.py" "<photo_folder>" [output_filename]
  • <photo_folder>: Path to folder containing year-named photos
  • output_filename (optional): defaults to photo_collage_年度照片.jpg

The script will:

  1. Auto-detect all photos (JPEG, PNG, HEIC) in the folder
  2. Extract the year from each filename (first 4 digits)
  3. Convert HEIC files temporarily using macOS sips
  4. Create a grid collage (5 columns) with dark year-label banners
  5. Save to the photo folder

Step 4: Open and Confirm

open "<photo_folder>/photo_collage_年度照片.jpg"

Show the user the output path and confirm success.

Example Interactions

Basic collage generation:

/photo-year-collage

→ Uses current directory, generates collage

With rotation:

User: rotate 2014, 2020 clockwise then make collage

→ Rotate those years, then generate collage

Specific folder:

/photo-year-collage ~/Documents/visa-photos

→ Uses specified folder

Requirements

  • macOS: sips (built-in) for HEIC conversion and rotation
  • Python 3 with Pillow: pip install Pillow --break-system-packages
    • Script will prompt to install if missing

Output

A single high-quality JPEG (photo_collage_年度照片.jpg) placed in the photo folder, showing all photos in a 5-column grid, each labeled with its year in a dark banner.

Skill Structure

photo-year-collage/
├── SKILL.md          (this file)
└── scripts/
    └── create_collage.py