feishu-content-extraction

Feishu Content Extraction

When to use

User asks to extract, summarize, or process content from a Feishu/Hera blog article URL (e.g., https://www.feishu.cn/content/article/XXXXX).

Problem

Feishu pages are heavily JS-rendered. The browser_navigate tool will time out trying to load them.

Approach

Step 1: curl + extract from HTML

curl -s -L -A "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36" "FEISHU_URL" 2>/dev/null

The article content is embedded as JSON in window._defaultTemplateValue and window._templateValue in the HTML. Key fields:

• window._defaultTemplateValue — contains text, title, description, keywords in the top-level page template
• window._templateValue — contains richtext.content (JSON-escaped string with Quill delta ops) and articleTitle, description, keywords

Step 2: Parse the content

• The text field in _defaultTemplateValue often contains raw text directly usable for summary
• The richtext.content in _templateValue contains Quill delta format JSON — {ops: [...]} with rich text operations. For detailed extraction, parse these ops by looking at each insert field and combining with attributes (bold, heading, etc.)

Step 3: Summarize

Once you have the text content, extract key sections, headings, and important points. Present as structured summary.

Pitfalls

• Never use browser_navigate for Feishu /content/article/ URLs — it will timeout (JS bundle is ~1.2MB)
• curl response is large (~1.2MB) — use grep or python to extract the specific JSON sections
• The embedded JSON uses escaped quotes (\\\") — handle escaping carefully when parsing
• Some content uses Quill delta format (not plain markdown) — you'll need to interpret the ops for full fidelity