垂直字体转换：优化中文排版 - Openclaw Skills

什么是 垂直字体转换？

垂直字体转换是 Openclaw Skills 生态系统中的一个专门工具，专为需要将标准水平字体转换为垂直阅读格式的开发人员和排版师设计。它优先考虑仅限本地的高精度工作流，确保标点符号、引号和混合脚本元素针对垂直布局进行正确定向。

该技能为源检查、分组字形转换和多阶段预览生成提供了一个强大的框架。通过利用复杂的字形分组规则而不是简单的旋转，它实现了专业级的结果，在保持原始字体美学完整性的同时，针对垂直阅读环境进行了优化。

下载入口:https://github.com/openclaw/skills/tree/main/skills/aster-copilot/ttf-converter

安装与下载

1. ClawHub CLI

从源直接安装技能的最快方式。

npx clawhub@latest install ttf-converter

2. 手动安装

将技能文件夹复制到以下位置之一

全局模式 ~/.openclaw/skills/ 工作区 <project>/skills/

优先级：工作区 > 本地 > 内置

3. 提示词安装

将此提示词复制到 OpenClaw 即可自动安装。

请帮我使用 Clawhub 安装 ttf-converter。如果尚未安装 Clawhub，请先安装（npm i -g clawhub）。

垂直字体转换 应用场景

• 为电子阅读器和移动应用创建专门的垂直阅读 TTF 字体。
• 修复现有字体中垂直标点、括号和省略号的方向问题。
• 生成本地审计产物和浏览器预览，以便在部署前验证字体渲染。
• 调整垂直中文文本流中拉丁字符和阿拉伯数字的位置。

垂直字体转换 工作原理

1. 该技能分析源水平 TTF，以确保其未损坏且适合转换。
2. 它生成初始水平浏览器预览，以建立质量和字符支持的基准。
3. 在用户确认后，它执行分组转换过程，将特定逻辑应用于不同的字形类别。
4. 渲染垂直浏览器预览，以便对垂直对齐和标点间距进行视觉检查。
5. 创建读者端 TXT 测试文件，以验证字体在真实阅读场景中的表现。
6. 在审核成功后，最终的垂直 TTF 和相关的测试产物将交付到本地。

垂直字体转换 配置指南

要将此功能集成到您的环境中，请确保已安装用于字体操作和渲染所需的 Python 依赖项。您可以使用以下命令运行完整的流水线：

# 在本地运行完整的转换流水线
python scripts/run_full_pipeline.py --input path/to/font.ttf --output_dir ./output

垂直字体转换 数据架构与分类体系

该技能将其输出组织到一个包含多个关键产物的结构化本地目录中：

文件类型	描述
.ttf	生成的垂直阅读 TrueType 字体文件。
.png	水平和垂直布局的基于浏览器的预览图像。
.txt	用于读者端验证的示例文件。
.json	包含字形边界框和中心数据的可选审计报告。

name: vertical-font-conversion
description: Local-only skill for converting horizontal Chinese fonts into high-quality vertical-reading TTF fonts with a preview-first workflow, grouped glyph rules, and local audit/test artifacts. Use when the user asks to make a 竖版字体 / 竖排 TTF, fix vertical punctuation, quotes, ellipsis, Latin, or digits, generate test images, or explicitly asks for Vertical Font Conversion.

Vertical Font Conversion

Use this skill when the task is about turning a horizontal font into a vertical-reading TTF font, or adjusting an existing vertical font result.

This skill is intentionally local-only. It focuses on source checking, grouped conversion, preview generation, local audit, and local test artifacts. It does not include remote conversion services, cloud uploads, or device-transfer steps.

Chinese display name

竖版字体转换

What this skill covers

• Source font precheck before vertical conversion
• Local vertical TTF production workflow
• Stable glyph-group handling rules
• Horizontal and vertical browser preview generation before delivery
• Reader test TXT generation
• Local audit for representative glyph placement
• Grouped conversion aimed at high-satisfaction vertical TTF output

Trigger conditions

Activate this skill when the user asks for any of the following:

• “制作竖版字体”
• “做竖排 ttf”
• “制作竖版 TTF”
• “改竖排标点 / 引号 / 省略号 / 西文 / 数字”
• “启动 Vertical Font Conversion”
• “按竖版字体转换流程来”

Default execution meaning

Unless the user explicitly asks to skip steps, “use Vertical Font Conversion” means:

• check the source font first
• make a horizontal preview first
• wait for user confirmation before building the vertical TTF
• make a vertical preview before delivering the font
• generate a reader test TXT
• stay fully local: horizontal preview, vertical TTF, vertical preview, reader test TXT, and audit output when needed

Mandatory workflow

Follow this order unless the user explicitly asks to skip steps:

1. Read the source horizontal TTF and check that the font opens normally and is not obviously broken.
2. Generate a browser test image from the original horizontal font first.
3. The preview must include:
   • Chinese body text
   • Latin letters
   • Arabic digits
   • single-point punctuation
   • book-title marks / quotes / brackets
   • dash / ellipsis
   • one mixed-script sentence
4. Ask the user to judge whether the source font itself is worth continuing.
5. If confirmed, make the vertical TTF by glyph groups.
6. Do not send the TTF immediately. Generate and show a browser test image for the vertical result first.
7. Also generate a TXT test file for reader-side validation.
8. Only after user confirmation, send the TTF and related local test artifacts.

If the original horizontal font already has obvious missing punctuation / Latin / digits / rendering defects, stop and tell the user it is not a good source candidate.

Read next

Before operating, read these references:

• references/workflow.md — full operating sequence
• references/rules.md — glyph-group rules and latest default parameters
• references/implementation-notes.md — practical logic, tuning philosophy, and failure-mode awareness for high-satisfaction output
• references/default-config.json — baseline tunable parameters
• references/examples.md — concrete command examples
• references/case-template.md — template for recording successful real cases
• references/cases-zihun-songkekaiti.md — confirmed real case and lessons from 字魂宋刻楷体

Script entry points

Prefer these bundled scripts before inventing fresh one-off code:

• scripts/render_original_preview.py — generate original horizontal preview image
• scripts/make_vertical_font.py — build glyf-based vertical TTF with the bundled grouped-conversion logic
• scripts/make_vertical_font_cff.py — build CFF-based vertical font with preserved CharString metadata
• scripts/render_vertical_preview.py — generate vertical result preview image
• scripts/generate_reader_test_txt.py — generate reader-side TXT sample
• scripts/audit_font_rules.py — inspect representative glyph bbox / center data after conversion
• scripts/run_full_pipeline.py — one-shot local pipeline for preview → build → preview → TXT

Use the bundled builders as the default baseline. If future tuning is needed, preserve the stable script paths and update the rule implementations behind them.

Scope boundary

This skill stops at local vertical-TTF deliverables.

Out of scope for this skill:

• epdfont conversion
• upload to /fonts
• remote processing services
• cloud transfer / device transfer steps

If a later workflow needs those actions, handle them in a separate skill or a separate explicitly user-directed procedure.

Operational rules

• Preserve the user-confirmed workflow; do not skip preview steps by default.
• Apply rules to the full font character set, not only the demo sentence.
• Process by glyph group; do not use one parameter set for all glyphs.
• For special symbols, prefer preprocessing in horizontal coordinates first, then produce the vertical result.
• Do not rely on only vert/vrt2 substitutions as the whole solution.
• Do not rotate first and then blindly patch positions.
• If the user provides a reference image, prefer matching the reference while keeping the same grouping logic.
• When a parameter tweak is needed, adjust the minimum necessary group rather than rewriting the whole rule set.
• Before sending a user-facing test image, do your own basic inspection first; do not throw raw debug output at the user as a substitute for checking.
• When the result still feels off, inspect representative glyph bbox / center data with the audit script instead of blind re-guessing.
• For one-shot execution, prefer the bundled full-pipeline script so the output order stays consistent.
• In the current WeChat direct channel, local file delivery should prefer the channel’s real file-send path with an absolute path rather than relying on MEDIA: lines alone.

Output expectations

Typical outputs include:

• horizontal preview image
• vertical preview image
• vertical .ttf
• test .txt
• optional audit JSON / terminal report