← Back to Skills Marketplace
kelcey2023

AI 书稿图片重设计

by kelcey2023 · GitHub ↗ · v1.0.99 · MIT-0
cross-platform ✓ Security Clean
261
Downloads
0
Stars
0
Active Installs
7
Versions
Install in OpenClaw
/install ai-book-image-redesign
Description
将用户提供的 HTML/前端原型整理成一个可复用的网页应用 skill。适用于用户说“把这段 HTML 做成 skill” “把这个网页原型封装成 skill”“生成一个可落地的静态页面 skill”“把图片重设计工具做成 skill”等场景。 该 skill 会把现成前端页面模板落到本地,默认输出为静态 HTM...
README (SKILL.md)

AI 书稿图片重设计 🖼️

把现成的网页原型(尤其是单文件 HTML/CSS/JS)整理成一个可复用 skill。

适用场景

当用户说这些话时使用:

  • 「把这段 HTML 做成 skill」
  • 「把这个网页原型封成 skill」
  • 「做一个静态网页工具 skill」
  • 「把这个 AI 图片重设计页面做成 skill」
  • 「把前端 demo 打包成可复用 skill」

默认工作方式

  1. 优先保留用户原始界面风格,不要擅自大改视觉结构
  2. 先做成本地可运行版本,再考虑发布或接 API
  3. 不在模板里写死密钥,改为让最终用户自行输入 API Base URL 和 API Key
  4. 静态资源放到 assets/,便于直接复制、部署、二次修改
  5. 如需脚本化初始化,用 scripts/ 放置生成脚本

目录说明

  • assets/index.html:可直接打开的静态页面模板
  • assets/vercel.json:Vercel 静态部署配置
  • scripts/setup.sh:把模板复制到目标目录,快速落地
  • scripts/export-static.sh:导出可直接部署到 Vercel / 静态站的文件

使用方式

1)快速落地页面

执行:

bash scripts/setup.sh /目标目录

默认会输出:

/目标目录/
└── index.html

1.1)导出可部署到 Vercel / 静态站的版本

执行:

bash scripts/export-static.sh /目标目录

会输出:

/目标目录/
├── index.html
└── vercel.json

部署方式:

  • Vercel:把该目录导入 Vercel,或在目录内执行 vercel
  • 静态站托管:直接上传 index.html(以及 vercel.json 可忽略)

2)页面中的 API 连接方式

当前模板改成由最终用户自行填写:

  • API Base URL
  • API Key
  • Model 名称
  • 可选的异步轮询路径(例如 /v1/tasks/{id}

填写后即可直接使用。页面默认请求:

POST {API_BASE_URL}/v1/images/generations

支持能力:

  1. 同步返回结果图
  2. 异步返回任务 ID 后自动轮询
  3. 多图批量处理(最多 10 张)
  4. 自定义 model 名称

并优先兼容这些返回字段:

  • data[0].url
  • data[0].b64_json
  • images[0].url
  • result.url
  • image_url
  • b64_json

任务 ID 兼容字段:

  • task_id
  • taskId
  • id
  • data.id

适合快速验证页面交互和图片生成流程。

但正式商用仍推荐:

  • 后端代理
  • 环境变量注入
  • 服务端签名中转

避免在浏览器前端长期直接暴露密钥。

修改建议

如果用户后续要求:

  • 批量处理多张图
  • 登录与额度系统
  • 支付升级
  • 服务端任务轮询
  • OSS / S3 图片存储

优先在现有模板基础上增量改,而不是推倒重写。

Usage Guidance
This appears coherent and low-risk for development/testing. Before installing or deploying: (1) review the full assets/index.html JavaScript to confirm there are no hidden external endpoints or telemetry; (2) never put a real API key into client-side HTML — follow the SKILL.md advice to use a backend proxy or server-side injection for secrets; (3) the provided scripts only copy files, so running them is safe, but check the exported files before deploying to production (especially if you plan to serve user-supplied images or billing features). If you want extra assurance, search index.html for any hardcoded URLs or tokens and remove or replace them.
Capability Analysis
Type: OpenClaw Skill Name: ai-book-image-redesign Version: 1.0.99 The skill bundle provides a template and utility scripts for an AI-powered image redesign tool. The core functionality is contained in a single-file HTML application (assets/index.html) that allows users to process images via an LLM API by providing their own API keys and endpoints (defaulting to api.apimart.ai). The shell scripts (scripts/setup.sh and scripts/export-static.sh) perform basic file management tasks, and the SKILL.md instructions correctly guide the AI agent to help users deploy the tool without any evidence of malicious intent, data exfiltration, or harmful prompt injection.
Capability Assessment
Purpose & Capability
Name/description say: convert a front-end prototype into a reusable static skill. The repo contains a static HTML template, a Vercel config, and two small shell scripts to copy or export those files — all expected for this purpose.
Instruction Scope
SKILL.md focuses on packaging the front-end, instructs running the provided setup/export scripts, and explicitly warns not to hard-code API keys and to prefer a backend proxy. The template loads Google Fonts and is designed to POST to a user-supplied API_BASE_URL; ensure you review the full index.html (including any JS) before deploying so you know which external endpoints will be contacted and that no analytics/telemetry endpoints are present.
Install Mechanism
No install spec; only small shell scripts that copy assets to a target directory. No downloads, no extracted archives, and no package manager installs — low-risk.
Credentials
The skill declares no required environment variables, no credentials, and no config paths. SKILL.md does discuss API Base URL / API Key as values the end user will supply in the UI, and scripts remind to replace placeholders — this matches the front-end nature of the skill.
Persistence & Privilege
always is false and the skill does not modify other skills or system settings. The provided scripts only copy files into a user-specified directory; they do not persist agent-wide configuration or request elevated privileges.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install ai-book-image-redesign
  3. After installation, invoke the skill by name or use /ai-book-image-redesign
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.99
- No user-visible changes in this version. - Documentation and file contents remain unchanged.
v2.0.0
支持自填URL+API Key+Model、批量多图、异步轮询、可部署Vercel
v1.0.4
Add Vercel/static-site deployable export
v1.0.3
Add async polling, batch processing, and custom model input
v1.0.2
Users can input API base URL and key, then use directly
v1.0.1
Let users input API base URL and API key
v1.0.0
Initial publish
Metadata
Slug ai-book-image-redesign
Version 1.0.99
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 7
Frequently Asked Questions

What is AI 书稿图片重设计?

将用户提供的 HTML/前端原型整理成一个可复用的网页应用 skill。适用于用户说“把这段 HTML 做成 skill” “把这个网页原型封装成 skill”“生成一个可落地的静态页面 skill”“把图片重设计工具做成 skill”等场景。 该 skill 会把现成前端页面模板落到本地,默认输出为静态 HTM... It is an AI Agent Skill for Claude Code / OpenClaw, with 261 downloads so far.

How do I install AI 书稿图片重设计?

Run "/install ai-book-image-redesign" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is AI 书稿图片重设计 free?

Yes, AI 书稿图片重设计 is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does AI 书稿图片重设计 support?

AI 书稿图片重设计 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created AI 书稿图片重设计?

It is built and maintained by kelcey2023 (@kelcey2023); the current version is v1.0.99.

💬 Comments