mirror of https://github.com/catlog22/Claude-Code-Workflow.git synced 2026-02-05 01:50:27 +08:00

Files

catlog22 169f218f7a feat(discovery): enhance discovery index reading and issue exporting

- Improved the reading of the discovery index by adding a fallback mechanism to scan directories for discovery folders if the index.json is invalid or missing.
- Added sorting of discoveries by creation time in descending order.
- Enhanced the `appendToIssuesJsonl` function to include deduplication logic for issues based on ID and source finding ID.
- Updated the discovery route handler to reflect the number of issues added and skipped during export.
- Introduced UI elements for selecting and deselecting findings in the dashboard.
- Added CSS styles for exported findings and action buttons.
- Implemented search functionality for filtering findings based on title, file, and description.
- Added internationalization support for new UI elements.
- Created scripts for automated API extraction from various project types, including FastAPI and TypeScript.
- Documented the API extraction process and library bundling instructions.

2025-12-28 19:27:34 +08:00

3.4 KiB

Raw Blame History

Phase 5: HTML Assembly

使用 universal-executor 子 Agent 生成最终 HTML，避免主 Agent 内存溢出。

核心原则

主 Agent 负责编排，子 Agent 负责繁重计算。

执行流程

const config = JSON.parse(Read(`${workDir}/manual-config.json`));

// 委托给 universal-executor 执行 HTML 组装
const result = Task({
  subagent_type: 'universal-executor',
  run_in_background: false,
  prompt: buildAssemblyPrompt(config, workDir)
});

const buildResult = JSON.parse(result);

Prompt 构建

function buildAssemblyPrompt(config, workDir) {
  return `
[ROLE] HTML Assembler

[TASK]
生成 TiddlyWiki 风格的交互式 HTML 手册（使用成熟库，无外部 CDN 依赖）

[INPUT]
- 模板: .claude/skills/software-manual/templates/tiddlywiki-shell.html
- CSS: .claude/skills/software-manual/templates/css/wiki-base.css, wiki-dark.css
- 配置: ${workDir}/manual-config.json
- 章节: ${workDir}/sections/section-*.md
- Agent 结果: ${workDir}/agent-results.json (含 tag 信息)
- 截图: ${workDir}/screenshots/

[LIBRARIES TO EMBED]
1. marked.js (v14+) - Markdown 转 HTML
   - 从 https://unpkg.com/marked/marked.min.js 获取内容内嵌
2. highlight.js (v11+) - 代码语法高亮
   - 核心 + 常用语言包 (js, ts, python, bash, json, yaml, html, css)
   - 使用 github-dark 主题

[STEPS]
1. 读取 HTML 模板和 CSS
2. 内嵌 marked.js 和 highlight.js 代码
3. 读取 agent-results.json 提取各章节 tag
4. 动态生成 {{TAG_BUTTONS_HTML}} (基于实际使用的 tags)
5. 逐个读取 section-*.md，使用 marked 转换为 HTML
6. 为代码块添加 data-language 属性和语法高亮
7. 处理 <!-- SCREENSHOT: id="..." --> 标记，嵌入 Base64 图片
8. 生成目录、搜索索引
9. 组装最终 HTML，写入 ${workDir}/${config.software.name}-使用手册.html

[CONTENT FORMATTING]
- 代码块: 深色背景 + 语言标签 + 语法高亮
- 表格: 蓝色表头 + 边框 + 悬停效果
- 内联代码: 红色高亮
- 列表: 有序/无序样式增强
- 左侧导航: 固定侧边栏 + TOC

[RETURN JSON]
{
  "status": "completed",
  "output_file": "${config.software.name}-使用手册.html",
  "file_size": "<size>",
  "sections_count": <n>,
  "tags_generated": [],
  "screenshots_embedded": <n>
}
`;
}

Agent 职责

读取模板 → HTML + CSS
转换章节 → Markdown → HTML tiddlers
嵌入截图 → Base64 编码
生成索引 → 搜索数据
组装输出 → 单文件 HTML

Markdown 转换规则

Agent 内部实现：

# H1 → <h1>
## H2 → <h2>
### H3 → <h3>
```code``` → <pre><code>
**bold** → <strong>
*italic* → <em>
[text](url) → <a href>
- item → <li>
<!-- SCREENSHOT: id="xxx" --> → <figure><img src="data:..."></figure>

Tiddler 结构

<article class="tiddler" id="tiddler-{name}" data-tags="..." data-difficulty="...">
  <header class="tiddler-header">
    <h2><button class="collapse-toggle">▼</button> {title}</h2>
    <div class="tiddler-meta">{badges}</div>
  </header>
  <div class="tiddler-content">{html}</div>
</article>

输出

{软件名}-使用手册.html - 最终 HTML
build-report.json - 构建报告

质量门禁

HTML 渲染正确
搜索功能可用
折叠/展开正常
主题切换持久化
截图显示正确
文件大小 < 10MB

下一阶段

→ Phase 6: Iterative Refinement

3.4 KiB Raw Blame History Unescape Escape