← 返回 Skills 市场
harrylabsj

Image Works

作者 haidong · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ 安全检测通过
32
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install image-works
功能描述
Batch image processing - compress, resize, format convert, watermark, EXIF clean, and crop. All local, no upload needed.
使用说明 (SKILL.md)

Image Workshop Skill

Local-first batch image processing toolkit. Compress, resize, convert formats, add watermarks, clean EXIF, crop — all on your machine with no image upload. Includes presets for Chinese social platforms (WeChat, Xiaohongshu, Taobao, etc.).

Core Capabilities

  • Compress: Quality target, size target, lossless, or auto-smart
  • Resize: Fixed dimensions, scale ratio, max-edge constraint, batch uniform
  • Convert: JPEG ↔ PNG ↔ WebP ↔ AVIF ↔ TIFF (bidirectional)
  • Watermark: Text or image, custom position/opacity, tile mode
  • EXIF: Remove all metadata, remove GPS only, or view metadata
  • Crop: Region crop, aspect-ratio crop (1:1, 4:3, 16:9, 3:4)
  • Presets: One-shot platform-optimized processing (WeChat, Xiaohongshu, Taobao, etc.)

Chinese Platform Presets

Platform Preset Name Specs
WeChat Moments wechat-moments 3×3 grid crop for 9-square layout
WeChat Official Account wechat-cover 900×383px (2.35:1), \x3C10MB
Xiaohongshu xiaohongshu 3:4 ratio, 1080×1440px
Taobao Main Image taobao-main 800×800px (1:1), \x3C500KB
Douyin Cover douyin-cover 1920×1080 (16:9)
Weibo Image weibo 1200px wide, \x3C20MB
Bilibili Cover bilibili-cover 16:9, 1920×1080
Avatar avatar 400×400 (1:1), \x3C200KB

Usage

clawhub run image-works --input \x3Cpaths> --op \x3Coperation> --output \x3Cdir>

Options

Option Type Default Description
--input json-array required File paths, directories, or glob patterns
--op json-array required Operations to apply (see below)
--preset string Platform preset (overrides --op)
--output.dir string ./processed/ Output directory
--output.suffix string '' Filename suffix
--output.overwrite bool false Overwrite existing files
--output.keep-structure bool true Preserve original dir structure
--recursive bool false Scan subdirectories
--max-files int 1000 Maximum files to process

Operations

Compress: {"type":"compress","quality":80,"target_size_kb":500} Resize: {"type":"resize","width":1200,"height":900,"fit":"inside"} Convert: {"type":"convert","format":"webp","quality":85} Watermark: {"type":"watermark","text":"© My Name","position":"bottom-right","opacity":0.5} EXIF: {"type":"exif","action":"remove"} Crop: {"type":"crop","aspect_ratio":"1:1"}

Operations are applied in array order.

Sample Prompts

1. Batch compress all JPEGs in a directory (most common)

clawhub run image-works --input './photos/*.jpg' \
  --op '[{"type":"compress","quality":80}]' \
  --output ./compressed/
# → 47 files processed, 156.2MB → 42.8MB (72.6% saved)

2. Xiaohongshu preset (one-shot)

clawhub run image-works --input ./product-photos/ \
  --preset xiaohongshu \
  --output ./xiaohongshu-ready/
# → Auto-resize to 3:4 (1080×1440), compress, export as JPEG

3. Add watermark with resize + format conversion

clawhub run image-works --input ./portfolio/ \
  --op '[{"type":"watermark","text":"© My Name","position":"bottom-right","opacity":0.3},{"type":"convert","format":"webp","quality":90}]' \
  --output ./watermarked-webp/
# → Watermarked + WebP converted

4. EXIF cleanup for privacy-safe sharing

clawhub run image-works --input ./vacation-photos/ \
  --op '[{"type":"exif","action":"remove"}]' \
  --output ./safe-to-share/
# → All GPS/device metadata removed

5. Batch resize for e-commerce (Taobao main image)

clawhub run image-works --input ./raw-products/ \
  --op '[{"type":"resize","width":800,"height":800,"fit":"cover"},{"type":"compress","quality":85}]' \
  --output ./taobao-ready/
# → Uniform 800×800 product images, optimized

First-Success Path

Step 1: Install → clawhub install image-works
Step 2: Run → clawhub run image-works --input ~/Pictures/ --op '{"type":"compress","quality":80}'
Step 3: See space-savings report → immediate value
Step 4: Explore presets → xiaohongshu, wechat-cover, etc.

Core Scripts

File Purpose
scripts/__init__.py Package init
scripts/scanner.py File scanner (glob, recursive)
scripts/processor.py Core processing pipeline
scripts/operations/compress.py Quality/target-size/lossless compression
scripts/operations/resize.py Dimension resize, fit modes
scripts/operations/convert.py Format conversion
scripts/operations/watermark.py Text/image/tile watermark
scripts/operations/exif.py EXIF read/remove/GPS-only
scripts/operations/crop.py Region/aspect-ratio crop
scripts/presets.py Platform preset definitions
scripts/reporter.py Processing report generator
scripts/progress.py Progress bar display
安全使用建议
Reasonable to install if you want local batch image processing. Review input globs and directory choices before running, because the skill can scan many local image files and write processed outputs; leave overwrite disabled unless you intentionally want existing output files replaced.
能力评估
Purpose & Capability
The stated purpose is local image compression, resizing, conversion, watermarking, EXIF cleanup, and cropping; the scripts implement those operations on user-supplied images.
Instruction Scope
Runtime scope is user-directed through input paths, operations, output directory, recursion, max-files, and overwrite options; recursion and overwrite are disabled by default.
Install Mechanism
Artifacts contain documentation, schemas, JSON presets, and Python processing scripts; no install hook, package-manager command, curl/wget fetch, or shell execution behavior was found.
Credentials
Local filesystem reads and output writes are proportionate for batch image processing; no network clients, subprocess execution, credential/session access, or unrelated data collection were found.
Persistence & Privilege
The skill creates output directories and processed image files, and can overwrite outputs only if the user enables that option. A user preset path constant is present but unused, with no background persistence or privilege escalation found.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install image-works
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /image-works 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
- Initial release of Image Workshop, a local-first batch image processing toolkit. - Supports compress, resize, format conversion, watermarking, EXIF cleaning, and cropping. - Includes ready-to-use presets for major Chinese social platforms (WeChat, Xiaohongshu, Taobao, etc.). - All processing is performed locally—no image upload required. - Flexible operation via command line with detailed options and operation chaining. - Provides user-friendly prompts, sample usage, and clear processing reports.
元数据
Slug image-works
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

Image Works 是什么?

Batch image processing - compress, resize, format convert, watermark, EXIF clean, and crop. All local, no upload needed. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 32 次。

如何安装 Image Works?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install image-works」即可一键安装,无需额外配置。

Image Works 是免费的吗?

是的,Image Works 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Image Works 支持哪些平台?

Image Works 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Image Works?

由 haidong(@harrylabsj)开发并维护,当前版本 v1.0.0。

💬 留言讨论