← Back to Skills Marketplace
guoqunabc

Feishu Readability

by Madoka · GitHub ↗ · v1.0.0
cross-platform ✓ Security Clean
292
Downloads
0
Stars
1
Active Installs
1
Versions
Install in OpenClaw
/install feishu-readability
Description
飞书文档格式优化与安全修改规范。**必须在通过MCP修改任何mi.feishu.cn文档之前加载此skill**。包括最小修改原则、富文本保护、段落间距、加粗规范。触发条件:用户要求修改飞书文档内容或格式,涉及mi.feishu.cn域名,或调用小米飞书MCP的update-doc工具。
README (SKILL.md)

飞书文档可读性优化 Skill

概述

本 Skill 旨在解决飞书云文档(Lark Doc)在自动化编辑中常见的格式问题,重点提升文档的视觉层次感和可读性。它基于对飞书 Markdown 渲染引擎底层逻辑(API vs. 交互式编辑)的深入理解,提供了一套经过验证的“最佳实践”。

🏆 三大核心原则

1. 🚀 核心铁律:最小修改原则 (The Principle of Minimal Modification)

权重:最高 (Critical) 定义:任何格式优化都必须是“手术刀式”的精准微调,严禁因为修改一个标点而重写整个段落,更不可重写整篇文档。

❌ 绝对禁止:

  • 严禁使用 overwrite 模式:除非旨在清空文档重写,否则绝不允许使用 overwrite 进行修改,这会导致评论、历史记录、图片 Token 和未知的富文本属性丢失。
  • 严禁大范围选区selection_with_ellipsis 不得跨越多个不相关的段落或标题。
  • 严禁盲目重构:不得在未获取该段落完整 Token(如图片、公式)的情况下重写包含该 Token 的段落。

✅ 最佳实践:

  • 字节级定位selection 应短至仅仅包含“待修改目标 + 确保唯一的最小上下文”。
    • ✅ 优秀:selection="token=\"xxx\"...**标题" (利用 Token 定位)
    • ✅ 优秀:selection="。**下一标题" (利用标点定位)
    • ❌ 错误:selection="上一段开头...下一段结尾" (范围太广,极易误伤)
  • 增量操作:优先使用 insert_before / insert_after 插入内容,而非 replace_range 替换内容。
  • 格式隔离:确保修改后的 Markdown 文本与原文在“未修改部分”完全一致(包括此时看不到的隐藏字符)。

2. 间距法则 (The Law of Visual Spacing)

权重:高 (High) 问题:API 写入的 Markdown 纯文本会被飞书解析器进行“空白清洗”(Whitespace Optimization),导致标准的空行被合并,文档显得拥挤。

✅ 解决方案:使用 带属性的占位段落 {align="center"}。 只有带有非默认属性(如 Center)的空段落,才会被解析器视为“有意义的内容”而强制保留。

📜 书写规范(基于最小修改原则):

  1. 插入式注入:使用 insert_before 在标题前插入 Spacer,而非 Replace 整个标题块。
  2. 必须以空格开头 {align...(无空格会被渲染为乱码文本)。
  3. 列表后阻断:若前文是列表,必须使用 双换行 \ \ 先结束列表,再插入标记。

3. 本土化格式卫生 (Formatting Hygiene)

权重:中 (Medium)

加粗规范

  • 冒号隔离:加粗必须止步于冒号之前,且冒号后必须有空格。
    • ✅ 正确:**标题**: 正文(推荐:冒号在加粗外,且有空格)
    • ✅ 可行:**标题:** 正文(若必须在内,冒号后必须分开)
    • ❌ 错误:**标题:**正文(粘连)

列表标点规范

  • 单句无标点:若列表项仅有一句话,去掉末尾的句号。
  • 富文本保护
  • 定位时务必避开 \x3Cimage>\x3Ccallout> 等标签。
  • 若必须修改包含标签的段落,需先 fetch-doc 获取 Token,并在 markdown 参数中原样写回。

🛠️ 标准操作流程 (SOP)

第一步:诊断 (Fetch & Analyze)

获取文档内容,重点检查:

  1. 标题前是否拥挤(特别是列表后的标题)。
  2. 加粗格式是否粘连。
  3. 是否存在特殊的富文本块需避开。

第二步:间距注入 (Spacer Injection)

使用 replace_range 在目标位置精确插入 Spacer。

场景 A:标准标题前(H1-H3)

mcp_FeishuMCP_update-doc(
    mode="replace_range",
    selection_with_ellipsis="前文结尾。\
## 标题", 
    markdown="前文结尾。\
\
 {align=\"center\"}\
\
## 标题"
)

场景 B:列表后的标题(高风险点)

mcp_FeishuMCP_update-doc(
    mode="replace_range",
    selection_with_ellipsis="- 列表项结尾\
## 标题",
    # 注意:双换行 + Spacer + 双换行
    markdown="- 列表项结尾\
\
 {align=\"center\"}\
\
## 标题"
)

第三步:格式精修 (Refinement)

修复粘连的加粗或排版错误。

mcp_FeishuMCP_update-doc(
    mode="replace_range",
    selection_with_ellipsis="**粘连标题**正文",
    markdown="**粘连标题:** 正文"
)

❓ 故障排除速查

现象 原因 解决方案
页面显示 {align="center"} 文字 缺少起始空格 替换为 {align="center"}
标题依然拥挤 列表后未双换行,或 Spacer 未独立占行 确保前后都有 `\
`
出现 width="100" 乱码 破坏了 \x3Cimage> 标签 Undo,重写时保留完整 Token
段落莫名消失/重复 replace_range 范围太大或定位不准 缩小定位范围,包含更多唯一的上下文标点

✅ 执行前核对清单 (Pre-flight Checklist)

  1. 定位准确性selection_with_ellipsis 是否足够短且唯一?
  2. 上下文安全:选区内是否包含了图片或引用块?(如有,需特别小心)
  3. Spacer 格式:是否检查了 {align="center"} 的空格和换行?
  4. 列表处理:前文如果是列表,是否加了 \ \
Usage Guidance
This skill appears coherent for its stated job: improving Feishu doc formatting with precise, token-aware edits. Before installing: (1) Confirm your agent/platform provides the MCP fetch-doc/update-doc tools and review the exact permissions those tools will have (limit to mi.feishu.cn if possible). (2) Test the skill on a copy of a document first — some examples use replace_range which can remove content if selection context is wrong. (3) Verify that image and other rich-text tokens are preserved by fetch-and-rewrite flows as described. (4) Ask the skill author to clarify the replace_range vs insert_before/after guidance to avoid accidental wide replacements. If you require stricter assurance, request an explicit permissions list for the MCP integration and a small test run log before using it on production documents.
Capability Analysis
Type: OpenClaw Skill Name: feishu-readability Version: 1.0.0 The skill bundle provides formatting guidelines and safety protocols for an AI agent interacting with Feishu (Lark) documents via a specific MCP tool (mcp_FeishuMCP_update-doc). It emphasizes 'minimal modification' to prevent data loss and provides specific workarounds for platform-specific Markdown rendering issues, such as using ' {align="center"}' for spacing. No evidence of data exfiltration, malicious execution, or harmful prompt injection was found; the instructions are focused on improving document readability and edit safety.
Capability Assessment
Purpose & Capability
Name/description state a narrowly scoped readability/formatting policy for mi.feishu.cn documents; the SKILL.md focuses on exact selection, spacers, and token preservation which matches that purpose. Minor inconsistency: the doc emphasizes a ‘minimal modification’ rule and preferring insert_before/insert_after, yet several concrete examples use replace_range — this is explainable but worth checking with the maintainer.
Instruction Scope
All runtime instructions stay within document editing scope: fetch-doc, analyze, and perform precise edits via the MCP update-doc calls. The skill explicitly warns to fetch tokens for paragraphs containing images/callouts and to preserve them — this is appropriate given the service’s tokenized rich-text model. It does not instruct reading unrelated files, system state, or exfiltrating data.
Install Mechanism
Instruction-only skill with no install spec and no code files — lowest-risk install posture. There are no external downloads, packages, or nonstandard installs.
Credentials
The skill declares no required environment variables or credentials, which matches the instruction-only nature. However, it assumes availability of MCP toolchains (mcp_FeishuMCP_update-doc, fetch-doc) and that those tools have the necessary permissions; the skill itself does not request or document the credential/permission scope it relies on, so verify that the platform-provided MCP integration will be limited to the intended mi.feishu.cn scope.
Persistence & Privilege
Skill is user-invocable and not always-on; it does not request elevated persistence or modify other skills or global agent settings. Autonomous invocation is allowed by default but not unusual; no 'always: true' or other high-privilege flags are present.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install feishu-readability
  3. After installation, invoke the skill by name or use /feishu-readability
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
飞书文档可读性优化 Skill 1.0.0 首次发布。 - 提供飞书文档格式优化最佳实践,强调最小修改原则,避免大范围或破坏性更新。 - 解决 API 写入导致的间距丢失问题,引入带属性空段落 `{align="center"}` 作为 Spacer 强制分隔段落。 - 明确加粗、列表等富文本处理规范,保护图片和特殊块的完整性,避免格式破坏。 - 给出标准操作流程(SOP),通过精准定位实现安全高效的自动格式优化。 - 故障排查与执行前清单,提升实用性和操作安全性。
Metadata
Slug feishu-readability
Version 1.0.0
License
All-time Installs 2
Active Installs 1
Total Versions 1
Frequently Asked Questions

What is Feishu Readability?

飞书文档格式优化与安全修改规范。**必须在通过MCP修改任何mi.feishu.cn文档之前加载此skill**。包括最小修改原则、富文本保护、段落间距、加粗规范。触发条件:用户要求修改飞书文档内容或格式,涉及mi.feishu.cn域名,或调用小米飞书MCP的update-doc工具。 It is an AI Agent Skill for Claude Code / OpenClaw, with 292 downloads so far.

How do I install Feishu Readability?

Run "/install feishu-readability" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Feishu Readability free?

Yes, Feishu Readability is completely free (open-source). You can download, install and use it at no cost.

Which platforms does Feishu Readability support?

Feishu Readability is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Feishu Readability?

It is built and maintained by Madoka (@guoqunabc); the current version is v1.0.0.

💬 Comments