lesson-builder
Class-meeting materials builder for university professors.
Skill by YujxZJCN
Chalk Labs
Need help deploying lesson-builder?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/yujxzjcn-lesson-builder
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: presentations, design, productivity, agent-building
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/lesson-builder
What's inside
- Skill name
lesson-builder- Description
- Class-meeting materials builder for university professors. 6-agent team turning a confirmed course design into lesson plans, lecture notes, slide outlines, active-learning activities, teaching cases, discussion guides, and flipped-classroom packages — one meeting or whole passport weeks at a time. Triggers on: lesson plan, lecture notes, slides, slide outline, in-class activity, active learning, case study, discussion questions, flipped classroom, prepare my class, teach next week, 备课, 教案, 讲稿, 课件, 课堂活动, 案例教学, 讨论课, 翻转课堂, 互动环节.
- License
- MIT
- Metadata
- status: active · version: 1.0.0 · last_updated: 2026-06-10 · pipeline_stage: 2
- Instructions
- 139 lines in SKILL.md · view on GitHub
Files (11)
- SKILL.md
- activity_designer_agent.md · agents/
- case_study_writer_agent.md · agents/
- discussion_designer_agent.md · agents/
- lecture_writer_agent.md · agents/
- lesson_planner_agent.md · agents/
- slide_outliner_agent.md · agents/
- active_learning_catalog.md · references/
- lesson_anatomy.md · references/
- activity_sheet_template.md · templates/
- lesson_plan_template.md · templates/
SKILL.md (first lines)
# Lesson Builder — Class-Meeting Materials Team
Builds the materials a professor actually walks into class with, from the design the
pipeline already confirmed: passport schedule weeks in, lesson packages out. The
professor brings the discipline content and knowledge of the room; this skill brings
lesson structure, active-learning craft, and tireless drafting — and it marks every
domain claim it cannot vouch for instead of sounding confident.
> **Prime rule:** a lesson serves the week's outcomes, not the other way around. If the
> professor asks for materials on a topic mapped to no outcome, build them — and flag the
> orphan once (Pedagogy Foundations §2), then let the professor decide.
## Quick Start
```
Build my Week 5 class on hash tables — 80 students, 75 minutes
帮我准备下周二的课,主题是回归分析,需要教案和课件大纲
I need an in-class activity for recursion that works with 120 students in fixed seating
Turn Week 9 into a flipped session: video spec + in-class problem session
Write a teaching case on a product-recall decision for my operations course
```
## Modes
| Mode | Trigger intent | Output |
|------|---------------|--------|
| `full` | "Prepare my class on X" / "build Week N" for one meeting | Complete package: lesson plan + lecture notes + slide outline + activity sheets |
| `lecture-notes` | "Write lecture notes / 讲稿 on…" | Full prose lecture notes with worked examples and speaker asides |
| `slides-outline` | "Make slides for…" | Slide-by-slide outline (assertion titles, figure specs) — not finished slides |
| `activities` | "I need an activity / make this interactive" | Active-learning activity design(s) with sheets and instructor logistics |
| `case-study` | "Write a teaching case on…" | Teaching case + teaching notes (discussion arc, board plan) |
| `discussion-guide` | "Discussion questions for…" / seminar prep | Question ladder + facilitation guide + async variant |
| `flipped` | "Flip this class" / pre-class video + in-class work | Pre-class material spec + in-class application session plan |
| `week-batch` | "Build Weeks 3–6" / pipeline Stage 2 run | `full` packages for each meeting in the range, one checkpoint per week |
**Mode dispatch rule:** ambiguous between a single artifact and `full` → ask which the
professor wants before building; a lesson package nobody asked for is wasted review
time. `week-batch` is the pipeline's mode — standalone professors usually want one
meeting at a time. Detect intent in any language.
### Does NOT trigger
| Scenario | Use instead |
|----------|-------------|
| Designing outcomes, the semester schedule, or the syllabus | `course-designer` |
| Writing graded instruments — exams, quizzes, rubrics, project briefs | `assessment-architect` |
| Full design → materials → assessment run across stages | `teaching-pipeline` |
(Ungraded in-class checks — minute papers, clicker questions used formatively — belong
here; the moment points attach, route to `assessment-architect`.)
## Agent Team (6)
| Agent | Role |
|-------|------|
| `lesson_planner_agent` | Designs the meeting arc: activation opener, timed segments, closure check, contingency notes; verifies the week's outcomes from the passport |
| `lecture_writer_agent` | Full prose lecture notes per segment: signal-before-detail, worked examples, anticipated misconceptions, `[VERIFY]` markers on uncertain claims |
| `slide_outliner_agent` | Slide-by-slide outlines: one idea per slide, assertion titles, figure specs with alt-text notes; flags where board work beats slides |
| `activity_designer_agent` | Selects and adapts techniques from `references/active_learning_catalog.md` to class size/modality; outputs activity sheets with full logistics |
| `case_study_writer_agent` | Teaching cases + teaching notes; synthetic scenarios are labeled, never passed off as real-world fact |
| `discussion_designer_agent` | Discussion guides: question ladders, equity-of-voice tactics, facilitation moves, online-async variant |
## Workflow (`full` mode)
```
Phase 0 LOAD — read the target week from course_passport.yaml: topic, outcomes,
assessments_due, learner_profile, class_size, modality, meeting
length. No passport → standalone intake: ask for exactly these
facts (Passport Iron Rule 2 — ask, don't guess), offer passport
creation at exit.
Phase 1 ARC — lesson_planner drafts the meeting arc per
references/lesson_anatomy.md: opener, segments with minute
timings, active-segment placement, closure, contingencies
🧑 checkpoint: arc confirmed (cheapest moment to change the lesson's shape —
present the timing budget, where actives land, and what got cut)
Phase 2 BUILD — parallel: lecture_writer (notes per segment) + slide_outliner
(outline keyed to segments) + activity_designer (sheets for each
active segment). All three build against the SAME confirmed arc;
none renegotiates timings unilaterally — conflicts go to the
checkpoint, not into the artifacts.
Phase 3 ASSEMBLE — package: lesson plan (templates/lesson_plan_template.md) +
lecture notes + slide outline + activity sheets; cross-check
segment IDs and timings agree across all four; collect every
[VERIFY] marker into one review list
🧑 checkpoint: package confirmed → passport week updated (activities[],
artifact_refs[], confirmed_by_professor)
```
Default lesson shape (overridable, surfaced as a "key decision" at the arc checkpoint):
at least one active segment per 20–25 minutes of lecture, scaled to class size and
modality (Pedagogy Foundations §4); worked examples before open problems when the
learner profile says novices (§9). `flipped` mode moves first-exposure input to the
pre-class spec and rebuilds the in-class arc around application. `week-batch` runs this
workflow per meeting; checkpoint cadence is per week, collapsing to minimal
confirmations when the professor says "just proceed" (Checkpoint Protocol).
## Iron rules
1. **Backward dependency.** Every lesson serves its week's outcomes from the passport.
Content serving no outcome is flagged at the arc checkpoint — once, with the orphaned
segment named — never silently built or silently dropped.
2. **No fabricated discipline content.** Facts, data, formulas, dates, and examples the
agent is not certain of carry `[VERIFY: <claim> — <why uncertain>]` inline. The
assembled package leads with the consolidated [VERIFY] list; a lesson ships with zero
unresolved markers or the professor's explicit acceptance.
3. **Activities are runnable, not aspirational.** Every activity states timing, group
size, materials, the launch script (what the instructor literally says), and failure
modes with fixes. "Have students discuss in groups" without logistics does not pass
Phase 3 assembly.
4. **UDL basics by default** (Quality Gate U1): heading structure in handouts, alt-text
notes on every figure spec, caption notes on any video spec, colorblind-safe palette
notes in graphics specs. Gaps are flagged at the package checkpoint, not left for
Gate 3.5 to find.
5. **Passport discipline.** After package confirmation, write `activities[]` and
`artifact_refs[]` to the week's schedule entry — append, never overwrite, and never
before the professor confirms (Passport Iron Rules 1, 4). Artifacts stay clean of
pedagogy citations; rationale at checkpoints cites Pedagogy Foundations §.
## OutputsFrom YujxZJCN/teaching-skills, MIT.
More from YujxZJCN/teaching-skills
- 0.571score
course-designer
YujxZJCN
Backward course design for university professors.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
accreditation-mapper
YujxZJCN
Accreditation and program-assessment documentation for university professors.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
bilingual-courseware
YujxZJCN
Bilingual course-material maintenance for university professors — default EN ⇄ 简体中文, language-pair agnostic.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
cohort-analyst
YujxZJCN
Learner and cohort analysis (学情分析) for university professors — cross-cutting support for design, builds, and the weekly loop.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
deck-studio
YujxZJCN
Slide deck production for university professors.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
lab-forge
YujxZJCN
Executable teaching-artifact builder for STEM professors — pipeline Stage 2 extension.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
media-scripter
YujxZJCN
Recorded-media scripting for university professors.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
student-mentor
YujxZJCN
During-semester student support for university professors.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
submission-auditor
YujxZJCN
Specification-driven submission auditing for university professors.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
ta-coordinator
YujxZJCN
Teaching-team management for university professors.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
teaching-pipeline
YujxZJCN
Full teaching-lifecycle orchestrator for university professors.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
teaching-reflector
YujxZJCN
Evidence-honest teaching reflection for university professors.
SkillCFreshness: C★ 41Claude Code +9
Signals
- Stars
- 41
- Forks
- 7
- Open issues
- 0
- Last commit (this folder)
- Jun 10, 2026 (4 months ago)
- Last release
- Jun 17, 2026 (4 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- Version
- 1.0.0
- License
- MIT
- Repository
- github.com/YujxZJCN/teaching-skills
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from YujxZJCN
- 0.571score
accreditation-mapper
YujxZJCN
Accreditation and program-assessment documentation for university professors.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
bilingual-courseware
YujxZJCN
Bilingual course-material maintenance for university professors — default EN ⇄ 简体中文, language-pair agnostic.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
cohort-analyst
YujxZJCN
Learner and cohort analysis (学情分析) for university professors — cross-cutting support for design, builds, and the weekly loop.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
deck-studio
YujxZJCN
Slide deck production for university professors.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
lab-forge
YujxZJCN
Executable teaching-artifact builder for STEM professors — pipeline Stage 2 extension.
SkillCFreshness: C★ 41Claude Code +9 - 0.571score
course-designer
YujxZJCN
Backward course design for university professors.
SkillCFreshness: C★ 41Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get lesson-builder running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
lesson-builder skill. Agent Mart. https://agentmart.dev/skills/yujxzjcn-lesson-builder (last updated 2026-10-24)