alioss-upload

Imported from wvlvik/toy-skills (skills/alioss-upload/SKILL.md). Install upstream with npx skills add wvlvik/toy-skills --skill alioss-upload. Copyright stays with the author.

OSS Upload

Upload files and directories to Alibaba Cloud Object Storage Service (OSS) using Python SDK v2.

Configure Environment Variables

Get Access Key

# 在 shell 配置文件中设置 (~/.zshrc, ~/.bashrc, 或 ~/.config/fish/config.fish)
export OSS_ACCESS_KEY_ID="your_access_key_id"
export OSS_ACCESS_KEY_SECRET="your_access_key_secret"
export OSS_REGION="cn-hangzhou"           # 可选, 默认: cn-hangzhou
export OSS_BUCKET="your-bucket-name"      # 可选, 也可通过 --bucket 参数传递
export OSS_DOMAIN="cdn.example.com"       # 可选, 自定义显示域名 (CDN 加速域名等)

# 使环境变量生效 (二选一)
source ~/.zshrc  # 或 source ~/.bashrc

Note: Environment variables must be set in your shell configuration. The script reads from os.environ, so it works automatically when variables are exported in your shell.

Output Behavior (Important)

After upload succeeds, display the OSS URL directly to the user!

When upload completes, output in this format:

✓ Upload complete!
  URL: https://bucket.oss-region.aliyuncs.com/path/to/file
  Key: path/to/file
  Size: 1.5 MB
  ETag: "abc123..."

For images uploaded with --image, the URL is publicly accessible.

Quick Start

Install Dependency

pip install alibabacloud-oss-v2

Important:

• Ensure the SDK is installed in the same Python environment that will run the script
• If you see "ModuleNotFoundError", check which Python is being used: which python3
• You may need to install to a specific Python: /Library/Frameworks/Python.framework/Versions/3.10/bin/python3 -m pip install alibabacloud-oss-v2

Upload Single File

# Basic upload
scripts/upload.py file.txt --bucket my-bucket

# Upload with custom key (path in bucket)
scripts/upload.py photo.jpg --key images/2024/photo.jpg --bucket my-bucket

# Upload with public-read access
scripts/upload.py document.pdf --bucket my-bucket --acl public-read

Upload Image (Public URL)

# Quick image upload with public-read ACL
scripts/upload.py photo.jpg --bucket my-bucket --image

# Output: https://my-bucket.oss-cn-hangzhou.aliyuncs.com/photo.jpg

Specify File Paths

When user wants to upload files, ask them to provide file paths in one of these ways:

1. Direct Path Input (推荐)

用户: 上传 /Users/me/Desktop/photo.png 到 statics/i/img
用户: 上传这张图片到 OSS: /path/to/image.jpg

2. Multiple Files

用户: 上传这些文件: /path/a.png, /path/b.jpg
用户: 批量上传 /Users/me/Downloads/*.png

3. Current File in Editor If user mentions "当前文件" or "this file", ask for the file path:

用户: 上传当前文件
助手: 请提供文件的完整路径，例如: /Users/me/Desktop/image.png

4. Drag & Drop Paths If user provides paths from drag-and-drop (common in terminal):

用户: 上传 /Users/me/Desktop/截图2024-01-01.png

Note: This model does not support image input directly. User must provide file paths as text, not attach images.

Upload from Clipboard (macOS)

使用剪贴板上传文件，无需手动输入路径：

# 步骤 1: 在 Finder 中选择文件
# 步骤 2: 复制文件 (Cmd+C)
# 步骤 3: 运行上传脚本

# 上传复制的文件为图片
scripts/upload-clipboard.sh --images

# 自定义前缀
scripts/upload-clipboard.sh --images --prefix statics/i/img/custom

# 指定 bucket
scripts/upload-clipboard.sh --images --bucket my-bucket

支持的剪贴板内容：

• Finder 中复制的文件 (Cmd+C)
• 复制的文件路径文本 (一行一个路径)

示例输出：

Files to upload:
  - /Users/me/Desktop/screenshot.png
  - /Users/me/Desktop/photo.jpg

✓ 上传完成: 2/2 张图片
  Bucket: my-bucket
  Region: cn-hangzhou
  Key 前缀: statics/i/img

Interactive Input (No Paths Required)

When no file paths are provided via command line, the script will prompt for input:

# Run without paths - will prompt interactively
scripts/upload.py --images --bucket my-bucket

# Interactive prompt example:
# Enter file paths to upload (one per line, empty line to finish):
#   Supported formats:
#     - Plain path: /Users/me/photo.jpg
#     - Path with spaces: /Users/me/My Photos/photo.jpg
#     - file:// URL: file:///Users/me/photo.jpg
#     - Home directory: ~/Desktop/photo.png
#   Or paste multiple paths (one per line):
# --------------------------------------------------
# /Users/me/Desktop/photo1.jpg
# /Users/me/My Photos/photo2.png
# ~/Downloads/screenshot.png
# file:///Users/me/Desktop/live-2746.png
#
# (press Enter on empty line to start upload)

Supported Input Formats:

• Plain paths: /Users/me/photo.jpg
• Paths with spaces: /Users/me/My Photos/photo.jpg
• file:// URLs: file:///Users/me/photo.jpg (automatically converted to local path)
• Home directory: ~/Desktop/photo.png (tilde expansion)
• Environment variables: $HOME/Desktop/photo.png (expanded automatically)

Path Validation:

• Invalid paths are automatically filtered out with warnings
• Duplicates are removed
• Paths are normalized and resolved to absolute paths

Stdin Pipe Input

# Pipe file paths via stdin
echo "/path/to/file1.jpg /path/to/file2.png" | scripts/upload.py --images --bucket my-bucket

# Or from a file containing paths
cat file_list.txt | scripts/upload.py --images --bucket my-bucket

Upload Multiple Images (推荐)

批量上传多张图片，自动处理中文/特殊字符文件名:

# 上传多张图片 (默认 key 前缀: statics/i/img)
scripts/upload.py img1.jpg img2.png photo.jpg --images --bucket my-bucket

# 自定义 key 前缀
scripts/upload.py photo1.jpg "照片.jpg" banner.png --images --bucket my-bucket --prefix assets/img

# 输出格式:
# ✓ 上传完成: 3/3 张图片
#   Key 前缀: statics/i/img
#
# 上传结果:
#   - statics/i/img/photo1.jpg (photo1.jpg)
#   - statics/i/img/a1b2c3d4e5f6.jpg (照片.jpg)  # 自动重命名
#   - statics/i/img/banner.png (banner.png)

特性:

• 中文/特殊字符文件名自动重命名为安全的 MD5 哈希名
• 保持原始文件扩展名
• 默认公开访问 (public-read)
• 返回 key + 原始文件名 格式

Upload Directory

# Upload all files in directory
scripts/upload.py ./images --prefix assets/images --bucket my-bucket --dir

# Upload with sanitized filenames (auto-rename Chinese/special chars)
scripts/upload.py ./photos --prefix photos --bucket my-bucket --dir --sanitize

# Upload only JPG files
scripts/upload.py ./photos --prefix photos --bucket my-bucket --dir --pattern "*.jpg"

# Non-recursive (skip subdirectories)
scripts/upload.py ./assets --prefix assets --bucket my-bucket --dir --no-recursive

特性 (with --sanitize):

• 中文/特殊字符文件名自动重命名为安全的 MD5 哈希名
• 保持原始文件扩展名
• 保持目录结构

Upload Large File (Resumable)

# Large file with automatic multipart and resumable upload
scripts/upload.py video.mp4 --bucket my-bucket --large

# Custom part size and parallelism
scripts/upload.py big-data.zip --bucket my-bucket --large --part-size 10 --parallel 5

Command-Line Options

scripts/upload.py PATHS [OPTIONS]
Arguments:
  PATHS                 Local file or directory paths to upload (supports multiple)
Required (one of):
  --bucket, -b          OSS bucket name (or set OSS_BUCKET env var)
  --prefix, -p          OSS key prefix (default for images: statics/i/img)
Upload Modes (mutually exclusive):
  --large, -l           Use multipart upload for large files (with resumable)
  --image, -i           Upload single image with public access
  --images              Upload multiple images (auto-sanitize filenames)
  --dir, -d             Upload directory contents

Options:
  --key, -k             OSS object key (default: filename)
  --region, -r          OSS region (default: cn-hangzhou)
  --domain, -D          Custom domain for display URLs (e.g., cdn.example.com)
  --acl                 Object ACL: private, public-read, public-read-write
  --storage-class       Storage class: Standard, IA, Archive
  --pattern             File pattern for directory upload (default: *)
  --no-recursive        Don't include subdirectories
  --sanitize, -s        Sanitize filenames with Chinese/special chars (for --dir)
  --part-size           Part size in MB for large files (default: 6)
  --parallel            Parallel uploads for large files (default: 3)
  --no-checkpoint       Disable resumable upload

Output:
  --quiet, -q           Suppress progress output
  --json                Output raw JSON response

Environment Variables:
  OSS_ACCESS_KEY_ID       Access Key ID (required)
  OSS_ACCESS_KEY_SECRET   Access Key Secret (required)
  OSS_REGION              Region (default: cn-hangzhou)
  OSS_BUCKET              Default bucket name
  OSS_DOMAIN              Custom domain for display URLs (e.g., cdn.example.com)
  OSS_CHECKPOINT_DIR      Checkpoint directory for resumable uploads

Upload Scenarios

1. Upload Website Assets

# Upload static assets to OSS
scripts/upload.py ./dist --prefix static --bucket my-web-bucket --dir

2. Upload Log Files to Archive

# Upload to Archive storage class for cost savings
scripts/upload.py logs.tar.gz --bucket archive-bucket --storage-class Archive

3. Upload User-Generated Images

# Batch upload with pattern
scripts/upload.py ./uploads --prefix user-content/2024 --bucket media-bucket --dir --pattern "*.jpg"

4. Upload Large Video Files

# Resumable upload for large files (handles network interruptions)
scripts/upload.py ./videos/webinar.mp4 --bucket video-bucket --large --key videos/2024/webinar.mp4

Error Handling

OSS URL Format

Uploaded files are accessible at:

https://{bucket}.oss-{region}.aliyuncs.com/{key}

Example:

https://my-bucket.oss-cn-hangzhou.aliyuncs.com/images/photo.jpg

Custom Domain (CDN)

If you have a custom domain or CDN bound to your OSS bucket, use --domain to display custom URLs:

# Using command-line option
scripts/upload.py photo.jpg --bucket my-bucket --image --domain cdn.example.com

# Or set environment variable
export OSS_DOMAIN="cdn.example.com"
scripts/upload.py photo.jpg --bucket my-bucket --image

Output will show:

✓ 上传完成!
  URL: https://cdn.example.com/statics/i/img/photo.jpg
  Key: statics/i/img/photo.jpg

Note: The custom domain only affects the displayed URL, not the actual upload process.

Storage Classes

Reference Documentation

• OSS Python SDK v2 Developer Guide
• OSS Console
• RAM Access Key Management
• OSS Regions and Endpoints