Read Aloud

Read documents or web content aloud via TTS.

When to use

When user says read this to me, read aloud, or listen to this article.

How to use

1. Identify the content source:
   • Feishu doc: feishu_doc_read to get content
   • Local file: read_file to get content
   • URL: fetch_url to get web content
2. Clean the text: strip HTML tags, markdown formatting, URLs
3. Split into paragraphs if long (TTS works best with shorter chunks)
4. Respond with the cleaned text — TTS will read it aloud
5. For very long content: summarize first, offer to read sections

Tip

Keep each response under 200 characters for natural TTS pacing. For longer content, break into multiple turns.

Example

User: "帮我朗读记忆文件" → read_file /data/agent/memory/MEMORY.md → Clean text, split into paragraphs → "你的记忆文件内容：用户偏好中文回复，常用嵌入式开发功能..." (TTS will read the response aloud)