Back to Details

k2f-cv

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

K2F Cv — AI-Native Aesthetic Design

K2F CV — AI 原生美学设计

Purpose

目的

Teach agents to blueprint CVs in 
content_and_design.md
: scholarly sections, tiered long lists, citation-ready entries, and design intent—without inventing degrees, papers, grants, or metrics.
指导 Agent 在 
content_and_design.md
中规划 CV：学术性章节、分层长列表、可直接引用的条目和设计意图——同时不得虚构学位、论文、资助或指标。

Built on K2F

基于 K2F 构建

This skill produces CVs as 
.K2F
packages. Follow the k2f skill for authoring, pack/verify, and export. K2F keeps agents on sections and list roles while the engine handles multi-page flow and typography.
Before you build, install the core skill: 
npx skills add suzheng/k2f --skill k2f
.
此 skill 会将 CV 生成 
.K2F
包。请遵循 k2f skill 进行编写、打包/验证和导出。K2F 让 Agent 专注于章节和列表角色，而引擎负责处理多页内容流与字体排印。
开始构建前，请安装核心 skill：
npx skills add suzheng/k2f --skill k2f
。

When to Use

使用场景

  • Building or updating an academic or research CV (faculty, postdoc, PhD, clinician-scientist, senior technical fellow).
  • Merging exports (Scholar, BibTeX, old Word) into one structure.
  • Selected vs full lists for a search—not one-page resumes (k2f-resume).
  • 构建或更新学术或研究型 CV（教职人员、博士后、PhD、临床科学家、高级技术研究员）。
  • 将导出内容（Scholar、BibTeX、旧版 Word）合并为统一结构。
  • 为岗位申请准备 Selected 与 Full 列表，而非一页式 resume（k2f-resume）。

Workflow

工作流

Step 1. Generate the content_and_design.md

第 1 步：生成 content_and_design.md

Single blueprint for semantic content and design intent. Use labeled blocks per section (identity, appointments, education, publications, etc.) that the k2f skill can map later.
Sub-steps
  1. CV frame — Discipline, career stage, opportunity type (tenure-track, fellowship, industry lab lead), geography, and page budget (often 2–6+ pages; justify length). Note required section order if the user cites a department or funder template.
  2. Input audit — Inventory appointments, degrees, publications, preprints, grants, teaching, service, talks, patents, supervision, awards, and links. Flag 
    NEEDS USER INPUT
    for missing DOIs, dates, or author order; never fabricate journals, amounts, or co-authors.
  3. Section architecture (include / cut / merge) — Default academic stack (reorder only when the frame demands it):
    • Identity & contact → Research interests (optional, short) → Education → Appointments / experience (reverse chronological) → Publications → Grants & funding → Teaching → Invited talks → Service & leadership → Awards → Skills / methods (sparse) → Supervision & mentoring.
    • Cut high-school, obsolete jobs unrelated to the search, duplicate talk lists, full grant narratives, hobby sections unless normative in field.
    • Merge thin consulting into appointments; split publications into Selected vs Full when lists exceed ~15–20 items or the user targets a page cap.
  4. Scan layers — Glance: name, affiliation, contact, brief interests. Scan: section titles and entry leaders. Detail: dates, venues, IDs—subordinate in design intent.
  5. Entry patterns — Publications: one style (authors, title, venue, year, DOI); peer-reviewed vs invited; own-name emphasis only if field-appropriate. Appointments: role, unit, dates; 0–3 impact lines with user-supplied metrics. Grants/teaching: agency, amounts, course + term when known.
  6. Tailoring pass — For a stated search, elevate matching themes (methods, disease area, pedagogy) via section order, a short interests line, and Selected publications—not by adding unverified work.
  7. Design intent — Tone, hierarchy, density for long documents:
    • Tone: scholarly, precise, calm; no marketing superlatives; active voice for impact lines, neutral for bibliography.
    • Hierarchy: name and current affiliation dominate; section titles clear and conventional; entry titles stronger than dates and parentheticals.
    • Density: list-forward; avoid essay blocks; publications as continuous list rhythm (hanging indent intent); whitespace between major sections, tight within homogeneous lists.
    • Palette: restrained academic; one accent; readable contrast in long lists.
Rules & constraints
  • Truth gate: every degree, title, paper, grant, and number traces to user input or an explicit stub.
  • Bibliography gate: do not invent DOIs, page numbers, or author lists; use 
    metadata incomplete
    rather than guessing.
  • Length gate: overflow → shorten via Selected lists and cut ancillary sections in the blueprint, not microscopic type.
  • Consistency gate: one date format, one citation style, stable institution names across sections.
  • Resume boundary: if the user needs a 1-page industry resume, redirect to k2f-resume in the blueprint note.
Common failure modes
  • Resume layout on a faculty CV → scholarly sections and publication emphasis.
  • Publication dump without citation rules → one style block for all entries.
  • Equal weight on every item → Selected plus optional full lists.
  • Missing appointment dates or degree years → halt with 
    NEEDS USER INPUT
    ; do not interpolate.
  • Keyword stuffing in a “Skills” wall → compact methods line or categorized skills tied to real work.
  • Inconsistent bolding of author name or mixed citation formats → single rule in design intent.
  • Burying current affiliation below ancient history → appointments and education ordered reverse-chron within each section.
Quality checklist
  • Frame: discipline, stage, opportunity, page budget, template constraints noted.
  • Section stack is conventional for field; cuts and Selected vs Full choices documented.
  • Glance layer shows identity, affiliation, and positioning without scrolling.
  • Publications (and grants/talks) share one citation pattern; own-name emphasis rule stated.
  • Appointments and education are complete on dates and institutions or flagged stubs.
  • Tailoring (if any) uses reorder/selection only—no fabricated entries.
  • Design intent covers tone, hierarchy, density, list rhythm, and accent for multi-page read.
使用单一蓝图同时表达语义内容和设计意图。按章节使用带标签的区块（身份、任职经历、教育经历、出版物等），供 k2f skill 后续映射。
子步骤
  1. CV 框架——学科、职业阶段、机会类型（tenure-track、fellowship、industry lab lead）、地域以及页数预算（通常为 2–6+ 页；需说明篇幅理由）。如果用户指定院系或资助方模板，记录所要求的章节顺序。
  2. 输入审查——清点任职经历、学位、出版物、预印本、资助、教学、专业服务、报告、专利、指导工作、奖项和链接。对于缺失的 DOI、日期或作者顺序，标记 
    NEEDS USER INPUT
    ；绝不虚构期刊名称、金额或共同作者。
  3. 章节架构（保留 / 删除 / 合并）——默认采用学术型章节结构（仅在框架要求时调整顺序）：
    • 身份与联系方式 → 研究兴趣（可选，简短）→ 教育经历 → 任职 / 工作经历（倒序）→ 出版物 → 资助与经费 → 教学 → 受邀报告 → 专业服务与领导力 → 奖项 → 技能 / 方法（精简）→ 指导与辅导。
    • 删除与目标岗位无关的高中经历、过时职位、重复的报告列表、完整的资助叙述以及兴趣板块（除非所在领域的惯例要求保留）。
    • 合并内容较薄的咨询工作与任职经历；当列表超过约 15–20 项，或用户设定了页数上限时，将出版物拆分为 Selected（精选）与 Full（完整）列表。
  4. 浏览层级——快速浏览（Glance）：姓名、所属机构、联系方式、简要研究兴趣。扫描（Scan）：章节标题和条目引导信息。细节（Detail）：日期、发表载体（venue）、ID——在设计意图中作为次要信息。
  5. 条目模式——出版物：统一采用一种格式（作者、标题、venue、年份、DOI）；区分同行评审与受邀文章；仅在符合领域惯例时突出本人姓名。任职经历：职位、机构、日期；附 0–3 行影响力说明，并仅使用用户提供的指标。资助 / 教学：在已知情况下填写资助机构、金额、课程和学期。
  6. 针对性调整——针对明确的目标岗位，通过章节顺序、简短的研究兴趣以及精选出版物来突出匹配主题（方法、疾病领域、教学法），而不是添加未经核实的工作。
  7. 设计意图——为长文档设定基调、层级和信息密度：
    • 基调：学术、精确、沉稳；不使用营销式最高级表述；影响力说明采用主动语态，参考文献信息采用中性语气。
    • 层级：突出姓名和当前所属机构；章节标题应清晰且符合惯例；条目标题的视觉权重高于日期和括号信息。
    • 密度：以列表为主；避免长篇段落；出版物保持连续列表节奏（体现悬挂缩进意图）；主要章节之间留白，同类列表内部保持紧凑。
    • 配色：采用克制、具学术感的配色；仅使用一种强调色；确保长列表中的对比度易于阅读。
规则与约束
  • 真实性门槛：每一项学位、职称、论文、资助和数字均可追溯到用户输入或明确设置的占位项。
  • 文献信息门槛：不得虚构 DOI、页码或作者列表；使用 
    metadata incomplete
    ，而不是进行猜测。
  • 篇幅门槛：内容溢出 → 在蓝图中通过 Selected 列表和删除辅助章节来缩短篇幅，而不是使用过小字号。
  • 一致性门槛：统一日期格式和引用风格；各章节中的机构名称保持一致。
  • Resume 边界：如果用户需要一页式 industry resume，请在蓝图备注中引导至 k2f-resume。
常见失败模式
  • 在 faculty CV 中采用 resume 布局 → 应采用学术章节并突出出版物。
  • 直接堆砌出版物且没有引用规则 → 所有条目应采用统一的格式块。
  • 所有条目的视觉权重相同 → 使用 Selected 列表，并视需要附加完整列表。
  • 缺少任职日期或学位年份 → 停止处理并标记 
    NEEDS USER INPUT
    ；不得推算。
  • 将“技能”堆成关键词墙 → 改用紧凑的方法陈述，或与真实工作挂钩的分类技能。
  • 作者姓名加粗规则不一致或引用格式混用 → 在设计意图中规定单一规则。
  • 将当前所属机构埋在久远经历之后 → 各章节内的任职与教育经历均按时间倒序排列。
质量检查清单
  • 框架：已记录学科、职业阶段、目标机会、页数预算和模板约束。
  • 章节结构符合所在领域惯例；已记录删除项以及 Selected 与 Full 的选择。
  • Glance 层无需滚动即可呈现身份、所属机构和定位。
  • 出版物（以及资助/报告）采用统一的引用模式；已说明本人姓名强调规则。
  • 任职经历和教育经历的日期与机构信息完整，或已标记为占位项。
  • 针对性调整仅使用重排和选择——没有虚构条目。
  • 设计意图涵盖基调、层级、密度、列表节奏和多页阅读所需的强调色。

Step 2. Use the content_and_design.md and k2f skill to generate a K2F file and export to an appropriate file format.

第 2 步：使用 content_and_design.md 和 k2f skill 生成 K2F 文件，并导出为适当的文件格式

Follow k2f to instantiate sections and flowing lists from the blueprint, apply roles from design intent, verify, and export the submission-ready CV (typically PDF or DOCX).
遵循 k2f，根据蓝图实例化章节和流式列表，应用设计意图中的角色，完成验证并导出可直接提交的 CV（通常为 PDF 或 DOCX）。