波斯语简历 / CV 生成器

安装指南

1. ClawHub CLI 从源直接安装技能的最快方式。
   npx clawhub@latest install persian-resume
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
   /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 persian-resume。如果尚未安装 Clawhub，请先安装（npm i -g clawhub）。
   更倾向于下载？
   获取包含原生技能文件的ZIP压缩包。
   下载技能 (.zip)

什么是

波斯语简历 / CV 生成器可将任何语言的职业信息或现有简历，转化为适用于实际求职的精致波斯语 CV。它支持 PDF 和 DOCX 输出，提供经典单栏和现代双栏模板，并根据候选人的实际经历调整各个部分，而不是强行加入无关内容。
与简单的翻译或排版工具不同，这项 Openclaw Skills 功能将简历视为双向文本文档：波斯语文本遵循逻辑阅读顺序，同时保留英文技术术语、URL 和电子邮件地址。它注重量化成果、日期一致性、事实准确性、字体易读性，以及对渲染后文档的验证。

应用场景

根据以波斯语、英语或其他语言提供的职业信息，创建专业的波斯语 CV。
在保留事实含义的前提下，改写或润色现有的波斯语简历。
修复包含英文职位名称、技术、URL 或电子邮件地址的简历中的 RTL 和双向文本问题。
生成用于求职申请的 PDF 或 DOCX 简历。
根据指定目标职位调整真实经历的重点和顺序，不虚构资历。
根据候选人资料和内容篇幅选择经典或现代版式。

工作原理

收集候选人的职业详情、目标职位、首选版式、日历、输出格式和模板偏好。
选择合适的结构和模板：传统单栏 CV 选择经典版，带侧边栏的双栏版式选择现代版。
按逻辑阅读顺序整理内容，优先呈现身份信息、近期经历、相关技能、成就和教育背景。保留拉丁字母技术术语和联系信息。
撰写简洁、以成就为导向的内容。仅在提供相关信息时使用可衡量的成果，绝不虚构缺失的事实、日期、公司、技术或指标。
使用 scripts/generate_resume.py 和简历 JSON 数据生成所需的 PDF 或 DOCX。
使用 scripts/validate_resume.py 验证双向文本、内容和文件完整性。
检查渲染后的输出，并在交付前修正拉丁文字错乱、日期不一致、字体不受支持或分页不当等问题。

配置指南

将波斯语简历 / CV 生成器技能及其 scripts/ 和 references/ 文件提供给 Openclaw Skills 环境。
收集经过核实的候选人职业信息和偏好。为简历选择一种日历系统；伊朗候选人优先使用波斯历，国际职位则适合使用公历。
按照 references/resume-data-schema.md 将简历输入整理为 JSON。通过 options.template 选择 classic 或 modern 模板。
确保系统提供支持波斯语的字体，例如 Vazir、Vazirmatn、Sahel、Shabnam 或 Samim。使用一致的字体栈，并保留英文技术术语的标准拼写。
使用提供的脚本生成并验证输出。技能文档未说明命令行参数，因此请查阅脚本或随附文档，确认确切的调用方式。
python scripts/generate_resume.py
python scripts/validate_resume.py
检查渲染后的 PDF 或 DOCX，而不只是源 JSON，并确认 RTL 流向、英文文本、联系信息、日期和分页均正确。

数据架构

该技能以 JSON 格式接收简历信息，输入结构记录在 references/resume-data-schema.md 中。使用 options.template（classic 或 modern）选择模板。文档列出了以下支持文件：
路径 用途
scripts/generate_resume.py 根据 JSON 数据生成 PDF 或 DOCX 简历。
scripts/validate_resume.py 检查 RTL 行为、内容和文件完整性。
references/resume-data-schema.md 定义简历输入 JSON schema。
references/validation-rules.md 说明详细的验证检查项。
相关简历内容可包括身份信息和职业头衔、联系信息、简介、工作经历、项目、教育背景、技能、证书、语言和奖项。职业经历条目可包含公司、职位、雇佣类型、地点、日期、成就要点和技术。日期应使用同一种日历并保持一致，同时保留拉丁字母 URL、电子邮件地址和技术术语。生成的交付文件为 PDF 或 DOCX；判断方向和版式时，应以渲染后的文件为准。

高级特性

处理波斯语和英语混合内容的双向文本，无需手动反转字符串，也不依赖单纯的右对齐。
生成 PDF 和 DOCX 后，专门验证 RTL 行为、内容和文件完整性。
提供两种模板：经典单栏版和带侧边栏的现代双栏版。
根据职业年限、内容密度和单页目标提供版式建议，同时避免不自然地压缩简历。
以成就为重点撰写内容，突出行动和成果，同时禁止虚构指标或经历。
根据目标职位优先展示经核实的相关经历，不伪造资历。
支持多页文档，保持页眉一致、页码齐全、章节衔接连贯，并留意不当分页。
采用便于 ATS 识别的结构，设置清晰标题和逻辑阅读顺序，并提供字体、日期、链接、混合方向文本和分页的质量检查清单。

name: persian-resume description: >- Generate professional Persian (RTL) resumes/CVs that render correctly with mixed
Persian-English content. Use this skill whenever the user asks to create,
rewrite, polish, or transform a Persian resume/CV, or when a resume contains
mixed Persian and English/technical content that must remain readable and
correctly ordered in RTL. Also trigger when the user provides career info in
any language and wants a professional Persian CV, or asks to fix RTL/bidi
issues in a Persian resume. This skill is NOT for general Persian documents —
use persian-documents for reports, letters, articles, and technical docs.
version: 0.1.0

Persian Resume / CV Generator

Turn career information into a polished, professional Persian RTL resume or CV that looks typeset, not generated. Supports both creating from raw info and transforming existing content.

When to use this skill

User asks to write, rewrite, or polish a Persian resume/CV
Resume mixes Persian and English (tech terms, company names, URLs)
User provides career info in any language and wants a Persian CV

Fixing RTL/bidi issues in an existing Persian resume

Output must be PDF or DOCX for real job applications

Not for: reports, letters, articles — use persian-documents instead

Core principle

A resume is a bidirectional document, not Persian text in a layout. The rendered output — not the source text — is the judge of correctness.

Never:

Reverse strings manually to "fix" RTL

Rely on right-alignment alone for RTL

Assume Unicode text guarantees correct rendering

Insert arbitrary ZWNJ/spaces as visual hacks

Always:

Use logical text order (reading order), let the renderer handle display

Keep English/Latin runs intact inside RTL paragraphs

Preserve English technical terms in their standard spelling

Validate the rendered output, not just the source text

Make the PDF/DOCX the source of truth for directionality

Resume structure

Choose sections based on the candidate's actual info. Do not force irrelevant sections.

[Name + Professional title]
[Contact row — RTL, with LTR URLs/emails intact]
[Professional summary — optional, 2-4 lines max]
[Experience]
[Company — Role — Dates]
• Achievement/outcome (not responsibility list)
[Projects] (if relevant to target role)
[Education]
[Skills] (structured categories, not a dump)
[Certifications / Languages / Awards] (only if present and meaningful)

Information hierarchy

Identity + professional positioning (name, title, summary)

Current/recent experience

Strongest relevant skills

Major achievements with outcomes

Education + additional info

Career history rules

Per entry:

Company name, job title, employment type, location
Start date, end date (or "current") 2-5 bullets: responsibility → action → outcome → metric

Technologies used

Dates — keep one calendar throughout:

Persian calendar preferred for Iranian candidates (۱۴۰۲–۱۴۰۵, ۱۳۹۹–اکنون)
Gregorian acceptable for international roles (2024–2026, 2021–Present)
Mixed formats: pick one and normalize, note the choice

Achievement-oriented writing

Prefer: action + measurable outcome over responsibility lists.

Good: راه‌اندازی سرویس → کاهش زمان پاسخ ۸۰۰ms به ۱۲۰ms (۸۵٪) Avoid: مسئولیت مدیریت تیم و نگهداری سرویس‌ها

Never invent metrics, companies, technologies, dates, or responsibilities. When info is missing, preserve known facts and leave gaps rather than fill them.

Professional summary

2-4 lines reflecting real background, communicating professional positioning, tailored to the target role when provided. Avoid generic buzzwords ("团队玩家", "چسبنده", "کار-hard"). Prioritize relevant expertise.

Skills structure

Group and prioritize — never an unorganized dump:

Programming languages

Frameworks & libraries

Frontend / Backend / DevOps (as relevant)

Databases & infrastructure

AI/ML (only if genuinely used)

Soft skills only when they add signal

Preserve standard English spelling for technical terms: JavaScript, TypeScript, React, Next.js, Django, FastAPI, PostgreSQL, Docker, Kubernetes, AWS, REST API, GraphQL. Do not transliterate.

Contact information

Clearly structured, visually separated from main content:

Phone, email, location (Persian)
Website, GitHub, LinkedIn, portfolio (LTR, intact)

URLs/emails must remain LTR and readable inside RTL

Templates

Choose a template via options.template in the resume JSON.

classic

Single-column layout

Traditional heading underlines

Contact info in a right-aligned strip below the name

Section titles with horizontal rules

Best for general-purpose professional use

modern

Two-column layout on A4

Right main column: summary, experience, education, projects
Left sidebar: contact, skills, languages, certifications

Compact sidebar with bold category titles

Best for tech/startup profiles, design-forward roles, or denser resumes

Layout & typography

Select layout based on content volume:

Single-column: < 8 years / compact profile
Two-column: 8-15 years with rich skills/projects
Compact: early-career / one-page target

Typography requirements:

Persian font: Vazir, Vazirmatn, Sahel, Shabnam, or Samim
Latin/technical: Arial, Noto Sans Mono for code
Body 10.5-12pt, headings 14-18pt
Line height 1.4-1.6, consistent spacing
Margins 1.5-2cm, balanced whitespace
Template consistency: use the same font stack across all sections in one template

Avoid: excessive decoration, oversized headings, dense unreadable blocks, tiny typography, confusing timelines, unbalanced whitespace.

Mixed Persian/English examples (must render correctly)
توسعه‌دهنده ارشد Frontend
توسعه و نگهداری سرویس‌های Django و FastAPI
Next.js 16
GitHub: github.com/ehsanghaffarii
email@example.com
کاهش هزینه ۳۰٪ با استفاده از React و Node.js

Multi-page rules

If experience requires 2+ pages:

Consistent header (name + page number) on each page

No orphaned headings at page bottom

Coherent section flow, no awkward splits

Do not compress to force one page

Content preservation (transform mode)

When transforming an existing resume:

Preserve all factual info; improve structure/wording without changing meaning
Remove duplication, fix consistency

Do not invent or silently remove experience

If user asks for shorter, prioritize relevant info — don't truncate arbitrarily

Target-role adaptation

When a target role is given, prioritize relevant experience/tech/achievements. Do not invent role-specific experience to fit the vacancy.

Output pipeline

Collect candidate info + target role + preferences (layout, calendar, format, template)
Select template: classic for traditional single-column, modern for two-column sidebar layout
Draft content in logical order (Persian text, English terms intact)
Generate via scripts/generate_resume.py → PDF or DOCX
Validate via scripts/validate_resume.py (bidi, content, rendering)
Render final output; re-check if any issue

Scripts

scripts/generate_resume.py — Generate resume from JSON data → PDF/DOCX
scripts/validate_resume.py — Validate RTL, content, file integrity
references/resume-data-schema.md — JSON schema for resume input data
references/validation-rules.md — Detailed validation rules and checks

Quality checklist

Before finalizing, verify:

RTL layout throughout (not just right-aligned)

English words/terms not reversed or scrambled

URLs, emails, file paths intact and readable
Version numbers, IDs, technical values unchanged
Dates consistent (one calendar, correct ranges)
Table column order correct (RTL)

Contact info with LTR content visually correct

Fonts support Persian glyphs

ZWNJ usage preserved where meaningful

Page breaks don't split content awkwardly

Mixed-direction sentences read naturally

ATS-friendly structure (clear headings, logical reading order)
Factual — no invented companies, metrics, dates, or roles
knownAs
操作公司邮箱