suzheng/k2f

k2f-worksheet

Design printable, fill-in-ready worksheets—practice sets, graphic organizers, lab records, reflections, or gap-closure forms—from learning goals and source material.

查看源码
仓库原始内容

按源仓库内容呈现,保留标题、案例、代码、表格、链接以及原文引用的演示图片。

K2F Worksheet — AI-Native Aesthetic Design

Purpose

Teach agents to blueprint a worksheet before layout: a single learner-facing task surface with scannable instructions, well-scoped items, and intentional response space—without essay dumps, duplicate prompts, or items that cannot be answered from what the sheet provides.

Built on K2F

This skill produces worksheets as .K2F packages. Follow the k2f skill for authoring: unpack or init a workspace, edit the semantic tree, run pack/verify, then export. K2F keeps agents on meaning (sections, items, response types)—not pixel coordinates—while the engine compiles layout, typography, and theme.

Before you build, install the core skill: npx skills add suzheng/k2f --skill k2f.

When to Use

  • Educators need practice, rehearsal, or application aligned to a lesson, unit, or standard.
  • The user wants structured blanks, organizers, short constructed responses, or checklists—not a full lesson plan (k2f-lesson-plan) or course overview (k2f-syllabus).
  • Analyst-style gap worksheets: turn messy inputs into categorized, answerable question blocks (inspired by gap-analysis patterns)—still learner- or facilitator-completable on paper.

Workflow

Step 1. Generate the contentanddesign.md

One blueprint for what appears on the sheet and how it should scan when printed. Register shared facts once; reference them from every item so the k2f skill can map sections without silent drift.

Sub-steps

  1. Route & clarify — Infer subject, grade band, duration, and worksheet type (drill, scaffolded practice, lab/data, reflection, organizer, gap-closure). Ask at most two high-leverage questions (e.g., target standard, with/without answer key); default item count and difficulty from context.
  1. Learning frame — Title, optional standard or skill gist (≤10 words), and what success looks like in one sentence. State prerequisites or materials the sheet assumes (textbook page, prior lesson, calculator)—not repeated per item.
  1. Information architecture — Layer top to bottom:
  • Orientation (brief): purpose and how to use the sheet; no teaching lecture.
  • Optional anchor (one block max): stimulus—passage excerpt, diagram description, data table, scenario—only if items depend on it.
  • Sections: grouped by sub-skill or cognitive step; each section has a short heading and 3–12 items unless the user specifies otherwise.
  • Response contract: per item, declare response type (blank line, checkbox, table cell, short paragraph box, label-on-diagram) and approximate space needed.
  • Answer key (if requested): separate outline keyed by item ID; do not interleave solutions with student items on the same page role.
  1. Item design — One measurable prompt per item; split compound questions. Order easy → hard within a section unless spiral review demands otherwise. For gap-style sheets, group by category; omit empty categories. Each gap row: short source snippet (or “not stated”), one specific question, blank answer field. Prioritize scope, core skill, and blocking ambiguities before nice-to-haves.
  1. Shared registry — Numbers, names, vocabulary, dataset values, and scenario facts live once under Shared content; items reference stable keys so edits cannot fork.
  1. Design intent — For the engine, not coordinates:
  • Tone: calm, classroom-credible; friendly for younger grades without clip-art clutter.
  • Hierarchy: title and learning frame strongest; section heads scannable; item numbers or letters consistent.
  • Density: instructions tight; generous workspace where handwriting or drawing is expected; avoid paragraph walls.
  • Consistency: parallel wording across items; same label scheme for sections and response types.
  • Palette: print-safe B&W; at most one accent for section headers or the learning frame.

Rules & constraints

  • Sheet must stand alone — Every item answerable from orientation + anchor + shared registry; no “discuss with teacher” placeholders without a written product.
  • Golden thread — Learning frame → sections → items → response spaces; cut items that do not practice the stated skill.
  • Original prompts — Source material may inform scope; never copy publisher problem sets or copyrighted stimuli verbatim.
  • Honest scope — If time or page budget is tight, reduce items or narrow the frame; do not shrink response space to illegibility.
  • Facilitator vs learner — Teacher notes or scoring rubrics belong in the blueprint as separate artifacts, not mixed into student-facing item text.

Common failure modes

  • Items test recall of unstated context → add anchor block or shared registry entry.
  • One question bundles multiple skills → split items and reorder.
  • Student sheet includes answer key or lengthy teaching → move to separate outline role.
  • Every section repeats the same instructions → lift to orientation once.
  • Gap categories with no real gaps → skip the section entirely.
  • Vague prompts (“Explain more”) → rewrite to name evidence, length, or format expected.
  • Mismatched difficulty in one section → rebalance or regroup.

Quality checklist

  • [ ] Learning frame and every item align; cut orphans.
  • [ ] Sections omitted when empty; item IDs stable for answer key.
  • [ ] Shared registry lists all cross-item facts with keys.
  • [ ] Each item has a declared response type and adequate space.
  • [ ] Reading level fits grade; stimuli length justified.
  • [ ] Design intent states tone, hierarchy, density, consistency, palette.
  • [ ] No verbatim third-party item banks; gap questions specific and answerable.

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

Follow k2f to instantiate the worksheet from the blueprint, apply roles from design intent, verify, and export print-ready PDF or other deliverables.

来自同一仓库

更多 Skills

全部 Skills
suzheng
社区

k2f-syllabus

Draft instructor-ready course syllabi with aligned outcomes, a scannable schedule spine, grading architecture, and policy layers from prompts, outlines, or legacy syllabi. Use when the user asks for a syllabus, course outline, curriculum map, or term schedule for a class they are teaching or redesigning. Create syllabi with AI-native design and the highest aesthetic standards, powered by the K2F format and engine. Purpose-built for AI agents, K2F lets them focus on semantic content and design in

安装量
14
GitHub Stars
2
最近更新
10月1日
suzheng
社区

k2f-invitation

Turn event facts and tone notes into print or digital invitations with ceremonial copy, layered logistics, and suite-consistent visual intent—not posters, programs, or slide decks. Use when the user asks for a wedding invitation, party invite, gala card, save-the-date, RSVP insert, or mobile e-invite. Create invitations with AI-native design and the highest aesthetic standards, powered by the K2F format and engine. Purpose-built for AI agents, K2F lets them focus on semantic content and design i

安装量
14
GitHub Stars
2
最近更新
10月1日
suzheng
社区

k2f-receipt

Blueprint transaction proofs and period impact receipts from user-supplied payment details or structured summaries—so totals, references, and attribution stay honest and scannable for customers, finance, or managers. Use when the user asks for a receipt, payment confirmation, sales slip, donation acknowledgment, expense proof, usage impact receipt, or “show what was delivered.” Create receipts with AI-native design and the highest aesthetic standards, powered by the K2F format and engine. Purpos

安装量
13
GitHub Stars
2
最近更新
10月1日
suzheng
社区

k2f-research-poster

Transform research papers, manuscripts, and findings into conference-ready academic posters with viewing distance typography, balanced columns, and high visual density. Use when asked to create an academic poster, scientific poster, conference poster, or A0/A1 poster. Create research posters with AI-native design and the highest aesthetic standards, powered by the K2F format and engine. Purpose-built for AI agents, K2F lets them focus on semantic content and design intent while automatically han

安装量
13
GitHub Stars
2
最近更新
10月1日