k2f-company-deck
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseK2F Company Deck — AI-Native Aesthetic Design
K2F 公司演示文稿 — AI 原生美学设计
Purpose
用途
Teach agents to blueprint company decks before layout: a stakeholder-ready story that explains who the company is, what it delivers, and why it matters—without inventing funding, customers, metrics, or leadership the user did not supply.
指导 Agent 在排版前规划公司演示文稿:构建可直接面向相关方的叙事,解释公司是谁、提供什么以及为何重要——绝不虚构用户未提供的融资信息、客户、指标或管理层。
Built on K2F
基于 K2F
This skill produces company decks as packages. Follow the k2f skill for authoring: unpack or init a workspace, edit the semantic tree, pack/verify, and export. K2F keeps agents on meaning (slide roles, narrative, visual intent)—not coordinates—while the engine compiles layout, typography, and theme into export-ready output.
.K2FBefore you build, install the core skill: .
npx skills add suzheng/k2f --skill k2f此 skill 将公司演示文稿生成 包。创作时遵循 k2f skill:解包或初始化工作区,编辑语义树,执行打包/验证并导出。K2F 让 Agent 专注于意义(幻灯片角色、叙事、视觉意图),而非坐标;引擎则将布局、排版和主题编译为可直接导出的成品。
.K2F开始构建前,安装核心 skill:。
npx skills add suzheng/k2f --skill k2fWhen to Use
适用场景
- Introducing the company to partners, enterprise buyers, recruits, press, or board guests (no investment ask).
- Consolidating website copy and bios into one coherent overview.
- Refreshing a generic “about us” deck for skimmable positioning and contact in under ten minutes.
- 向合作伙伴、企业买家、招聘对象、媒体或董事会来宾介绍公司(不提出投资请求)。
- 将网站文案和简介整合为一份连贯的企业概览。
- 刷新通用的“关于我们”演示文稿,使其便于快速浏览、定位清晰,并能在十分钟内以联系信息收尾。
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 aesthetic intent. Use labeled blocks per slide (role, headline, body, visual type) that the k2f skill can map later.
Sub-steps
-
Deck frame — Primary audience (partner, customer, talent, media), setting (first meeting, conference, onboarding), and time budget. Choose emphasis: credibility + capabilities for B2B; mission + culture for talent; news + proof for press. Drop fundraising slides unless the user explicitly wants a hybrid.
-
Input audit — Inventory verified inputs: positioning, products, public pricing, permitted logos, milestones, team, partnerships, news. Flagfor gaps—never fabricate headcount, revenue, clients, or awards.
NEEDS USER INPUT -
Narrative spine — Plan flow before titles: identity → problem space → offer → how it works → proof → difference → people & scale → direction (vision, not roadmap dump) → engage. Opening = name + tagline; closing = one CTA.
-
Slide architecture (include / cut) — Default overview stack (~12–16 slides); merge or omit when inputs are thin:
- Title → Company snapshot → Mission / values (optional) → Problem & opportunity (light) → Solutions / product lines → How we work or platform → Customers & outcomes → Differentiation → Team → Milestones / footprint → Partners or ecosystem (optional) → Vision → Contact.
- Cut investor-only blocks (TAM essay, use-of-funds, cap table), duplicate product slides, long feature lists, and internal org charts unless audience is recruits.
- Split when identity and product story share one slide, or when two products each need a hero proof point.
-
Per-slide content contract — For each slide specify:
- Headline: outcome or position, not a section label (“Payments infrastructure for internet businesses” not “About us”).
- Body: max 3–5 bullets or one visual; ~2 lines per bullet; parallel grammar on lists.
- Visual type: logo strip, product frame, workflow diagram, metric callouts, timeline, team grid, map—prefer visual intent over paragraphs for comparisons, milestones, and portfolios.
- Footnotes: source + year for external claims; logo approval notes if needed.
-
Proof layering — Permitted logos or named references first; anonymize if required. Scale metrics only when user-supplied; else qualitative proof with TBD.
-
Design intent — For the engine: tone institutional, from user brand voice—not generic hype; hierarchy one hero per slide, calm contact block; density one message per slide, grids for portfolios; palette user colors or neutral + one accent, WCAG AA, consistent logos/timelines.
Rules & constraints
- Audience gate: every slide must earn its place for the chosen audience; recruit slides ≠ partner slides unless user wants one master deck.
- Truth gate: metrics, customers, funding, and headcount match user materials or are marked TBD—no rounded-up fiction.
- One-idea gate: if the headline needs “and,” split or cut.
- Brand gate: tagline, product names, and legal entity spelling identical everywhere.
- Logo gate: only logos the user confirmed; otherwise “category examples” or omit.
Common failure modes
- Pitch deck as company deck → drop ask, financials, TAM; strengthen identity and contact.
- Website walls of text → headline + proof only; detail in blueprint notes.
- Feature lists without outcomes → tie capabilities to customer results.
- Weak differentiation → two-column comparison or cut.
- Long team bios → name, role, one line each.
- Cluttered closing → one CTA plus minimal contact.
Quality checklist
- Tagline and company snapshot appear in the first three slides; contact/CTA on the last slide.
- Every slide has one headline claim and a declared visual type.
- No slide exceeds density limits; external stats have sources or TBD flags.
- Product and company names spelled consistently; repeated metrics match.
- Audience-irrelevant sections removed or moved to optional appendix notes in the blueprint.
- Design intent names tone, hero emphasis, and accent usage for the full deck.
为语义内容和美学意图生成一份统一蓝图。对每张幻灯片使用带标签的内容块(角色、标题、正文、视觉类型),供 k2f skill 后续映射。
子步骤
-
演示框架 — 确定主要受众(合作伙伴、客户、人才、媒体)、应用场景(首次会面、会议、 onboarding)以及时间预算。选择侧重点:B2B 场景侧重可信度 + 能力;人才场景侧重使命 + 文化;媒体场景侧重新闻 + 佐证。除非用户明确需要混合版本,否则删去融资相关幻灯片。
-
输入审计 — 盘点已验证的输入材料:定位、产品、公开定价、获准使用的 logo、里程碑、团队、合作关系、新闻。对于缺失项标记——绝不虚构员工人数、营收、客户或奖项。
NEEDS USER INPUT -
叙事主线 — 先规划流程,再拟定标题:身份 → 问题领域 → 产品/服务 → 运作方式 → 佐证 → 差异化 → 团队与规模 → 发展方向(愿景,而非堆砌路线图)→ 促成行动。开篇 = 名称 + 宣传语;结尾 = 一个 CTA。
-
幻灯片架构(包含 / 删减) — 默认概览结构(约 12–16 页);当输入材料较少时,合并或省略:
- 标题 → 公司概览 → 使命/价值观(可选)→ 问题与机会(简要)→ 解决方案/产品线 → 工作方式或平台 → 客户与成果 → 差异化 → 团队 → 里程碑/业务布局 → 合作伙伴或生态系统(可选)→ 愿景 → 联系信息。
- 删减 仅供投资方使用的内容(TAM 长文、资金用途、股权结构表)、重复的产品页、冗长的功能列表和内部组织架构图;但如果受众是招聘对象,组织架构图可以保留。
- 拆分 当公司定位与产品故事共用一页时,或当两种产品各自需要一个主视觉佐证点时。
-
每页内容约定 — 为每张幻灯片明确以下内容:
- 标题:表达结果或定位,而不是栏目名称(例如使用“面向互联网企业的支付基础设施”,而不是“关于我们”)。
- 正文:最多 3–5 个要点或一个视觉元素;每个要点约 2 行;列表采用并列结构。
- 视觉类型:logo 组合条、产品界面框图、工作流图、指标标注、时间线、团队网格、地图——对于比较、里程碑和产品组合,优先表达视觉意图,而不是使用大段文字。
- 脚注:外部主张标注来源和年份;必要时注明 logo 审批说明。
-
分层构建佐证 — 优先使用获准的 logo 或具名引用;如需则匿名处理。仅在用户提供数据时展示量化指标;否则使用定性佐证并标注 TBD。
-
设计意图 — 为引擎明确以下规范:语气 应具备机构级专业感,遵循用户的品牌表达,而不是泛泛的炒作话术;层级 每页一个主视觉,联系信息块保持克制;密度 每页一个信息,产品组合使用网格;配色 使用用户提供的颜色,或中性色 + 单一强调色,符合 WCAG AA,logo 与时间线保持一致。
规则与约束
- 受众规则:每一页都必须对所选受众有存在价值;除非用户需要一份通用母版,否则招聘版页面 ≠ 合作伙伴版页面。
- 真实性规则:指标、客户、融资信息和员工人数必须与用户材料一致,否则标记为 TBD——不得虚构或使用向上取整的数据。
- 单一信息规则:如果标题需要用“和”连接,就拆分或删减。
- 品牌规则:宣传语、产品名称和法人实体拼写在所有页面中保持一致。
- Logo 规则:仅使用用户确认的 logo;否则使用“类别示例”或省略。
常见失败模式
- 把融资路演稿误作公司概览 → 删除融资诉求、财务数据和 TAM;强化公司定位与联系信息。
- 直接堆砌网站大段文字 → 仅保留标题和佐证;详细内容放入蓝图说明。
- 功能列表没有成果 → 将能力与客户结果关联。
- 差异化表达薄弱 → 使用双栏对比,或删减。
- 团队简介过长 → 仅保留姓名、职位和一行说明。
- 结尾信息杂乱 → 保留一个 CTA 和最精简的联系信息。
质量检查清单
- 宣传语和公司概览出现在前三页;联系信息/CTA 位于最后一页。
- 每页都有一个标题主张,并明确声明视觉类型。
- 任何页面的信息密度都不得超过限制;外部统计数据必须有来源或 TBD 标记。
- 产品和公司名称拼写一致;重复出现的指标保持一致。
- 与受众无关的章节已删除,或在蓝图中移至可选附录说明。
- 设计意图明确整套演示文稿的语气、主视觉强调方式和强调色用法。
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 slides from the blueprint, apply roles from design intent, verify, and export the stakeholder-ready deliverable.
遵循 k2f skill,根据蓝图实例化幻灯片,应用设计意图中的角色,进行验证并导出可直接面向相关方的交付物。