Plan Lesson

Plan Lesson

Think a single class session through using Understanding by Design (UbD) — start with what students should understand, design how you'll know they got there, then build the activities that get them there. Lesson-scaled: 1–2 learning targets, one formative check, one class arc. Produces a UbD brief in chat; offers handoff to post-lesson-page when the teacher's ready to materialize it on Canvas.

This skill writes nothing to Canvas. It's a pedagogical thinking partner. Posting is a separate skill (post-lesson-page) the teacher can run once the brief is locked.

When to use

A teacher is figuring out what to teach for a single class session and wants a framework. Common phrasings:

• "Plan a lesson on the water cycle for DSGN 9 — help me think through it"

• "I'm teaching prototyping tomorrow, not sure how to structure it"

• "Design a lesson on parallel circuits using UbD"

• "What's a good way to frame this watershed lesson?"

• "Help me think backwards from what I want students to understand"

When NOT to use

• The teacher already knows what they're teaching and just wants it on Canvas. That's post-lesson-page. Don't make them think backwards if they don't want to.

• Multi-class units or modules. That's plan-module (which has its own UbD path scaled to multi-lesson arcs).

• Assessments / tests / projects. That's plan-assessment for the page, post-quiz for the actual Canvas Quiz object.

• Whole-course scaffolding. That's plan-course.

Prerequisites

• canvas-mcp v0.3.16 or later installed if you want school defaults (class period length, current academic week) pulled via get_school_info. The skill works without canvas-mcp for pure ideation, but will have to ask the teacher for period length and module context every time.

• No Canvas writes happen in this skill. No published: false worry, no draft creation — this is a chat-only working document.

Workflow

1. Establish context

Confirm with the teacher (asking only for what's missing):

• Lesson topic / focus. What's the lesson about, in one phrase. If they gave a working title, use it; otherwise propose one.

• Class duration. First, call get_school_info once at the start of step 1. If academic_calendar.class_period_minutes is set, use it silently as the default (e.g., Franklin = 50 min). Only ask if the school config doesn't specify a period length AND the teacher didn't either.

• Where does this lesson sit? Standalone, or part of an existing Canvas module / unit plan? If part of a module, optionally call list_modules(course_identifier) + get_page_content on the module's outcomes page to pull the module's enduring understandings and essential questions — the lesson should advance those, not invent new ones.

• Which Canvas course? Only needed if you'll pull module context or call list_competencies. If the teacher is purely ideating without a course in mind, skip this — you don't need a course to think about a lesson.

Don't over-confirm. If the teacher said "plan a 60-min lesson on prototyping for DSGN 9," you have topic, duration, and course — move on.

2. Stage 1 — Desired Results

For a single lesson, scope is smaller than a module. Identify:

Learning targets (1–2 statements, SWBAT form — "Students will be able to…"). Specific, observable, achievable in one class. Distinguish for the teacher:

Essential question (1, sometimes 2). The open-ended question students wrestle with during this class. Often borrowed from the module's essential questions — pick the one most relevant to today's targets. If standalone (no module), draft one.

Enduring understanding(s) (1, sometimes 2). The big idea this lesson builds toward. Almost always borrowed from a parent module if there is one; if standalone, draft one but don't overreach — a single lesson advances an EU, it doesn't deliver it.

Competency alignment. Call list_competencies if a course is set and the school config has one. Pick 1–2 competencies this lesson naturally exercises through the targets + planned activity. Don't shoehorn — a vocab-memorization lesson doesn't target Systems Thinking just because the topic is ecosystems.

For conventional content (Algebra topics, US History units, standard Biology, world languages): draft these for the teacher to react to. Claude has reasonable pattern-matching here.

For unconventional or teacher-designed content (transdisciplinary projects, school-specific curriculum, novel framings): don't assume. Share initial ideas as conversation starters, let the teacher shape direction.

Stage 1 checkpoint — show the teacher:

## Stage 1: Desired Results

**Learning targets** (SWBAT):
1. [...]
2. [...]

**Essential question:** [...]

**Enduring understanding:** [...]

**Competency focus:** [Knowledge-Based Reasoning, Systems Thinking, ...]

Wait for the teacher to confirm or revise before moving to Stage 2. This is the foundation — getting it right matters more than speed.

3. Stage 2 — Evidence of Understanding

At the lesson scale, design one formative check that tells the teacher whether students hit the targets. The check should:

• Match the targets. If the target is "explain why endorheic basins don't drain to oceans," the check needs students producing an explanation — not multiple choice.

• Fit inside the class period. 3–10 minutes typically. Exit tickets, mini-demos, quick-writes, peer explanations, artifact checks, observation rubrics.

• Be specific. "I'll see if they got it" is not a formative check. "I'll collect a 2-sentence written response to 'Why does the Great Salt Lake have no outflow?' and look for mention of evaporation + closed-basin geography" is.

Format the formative check:

• What it is (exit ticket / quick-write / observed demo / mini-artifact / discussion prompt)

• When in the lesson (last 5 min? mid-class transition? continuous observation?)

• Success signal (what a student who hit the target produces vs. one who didn't)

If the lesson feeds a graded summative arc (unit test next week, project due Friday), note the connection but don't design the summative here — that's plan-assessment's job.

Stage 2 checkpoint — show the teacher:

## Stage 2: Evidence of Understanding

**Formative check:**
- Type: [exit ticket / quick-write / ...]
- When: [end of class / mid-class / ...]
- Prompt or task: [...]
- Success signal: [what a student who hit the target produces]
- Miss signal: [what a struggling student produces]

**Connection to summative** (if any): [...]

Wait for confirmation before Stage 3. If the formative check feels mismatched to the targets, push back — that's the most common UbD failure mode at the lesson scale.

4. Stage 3 — Learning Plan

Plan the arc of the class session — phase-by-phase activities that get students from where they walked in to where the targets land. Match the rough total to the duration confirmed in step 1.

Format as a timed sequence:

0:00–0:05  Hook: [what gets students into the question]
0:05–0:15  Activator: [activate prior knowledge / surface misconceptions]
0:15–0:35  Main investigation: [the heart of the lesson]
0:35–0:50  Consolidation: [synthesis, discussion, structured talk]
0:50–0:55  Formative check (from Stage 2)
0:55–1:00  Closing / preview next class

Per phase, capture:

• Name + minutes (so total adds up to the class length)

• What students are doing (active verbs — investigating, comparing, defending, building, sketching, arguing)

• What the teacher is doing (modeling, observing, questioning, scaffolding — not always lecturing)

• Materials / prompts / links (only if known — leave TBD otherwise)

Sequencing principles:

• Hook → exploration → consolidation → check → close. UbD prefers students encountering the question before receiving the answer.

• Active beats passive. If a phase has students "listening to the teacher explain X," ask whether they could discover X instead.

• Less direct instruction than feels natural. 10–15 min of teacher talk per 60 min is plenty for most secondary classes.

• Make space for student talk + iteration — turn-and-talks, peer feedback, draft-revise cycles.

• Don't pack the schedule. Buffer time matters; ambitious pacing that requires perfection will fail.

• The formative check from Stage 2 has a slot in the timeline — show it explicitly.

Stage 3 checkpoint — show the teacher the timed sequence. Wait for confirmation.

5. Compose the full UbD brief (when useful)

If the three Stage checkpoints are still clearly visible above in the chat, you can skip a redundant restatement and go directly to step 6 — the brief already exists in the conversation.

Compose a single block when either of these is true:

• The conversation has drifted (many turns since Stage 1, the teacher may not see it scrolling back), OR
• The teacher will likely hand off to post-lesson-page — having one copy-pastable artifact makes the handoff cleaner.

The composed form when you do present it:

# Lesson Plan — [Title]
*[Course] · [Date or Week] · [Duration] min*

## Stage 1: Desired Results
**Learning targets (SWBAT):**
1. [...]
2. [...]

**Essential question:** [...]

**Enduring understanding:** [...]

**Competency focus:** [...]

## Stage 2: Evidence of Understanding
**Formative check:** [type] — [prompt or task]
- When: [...]
- Success signal: [...]

## Stage 3: Learning Plan
| Time | Phase | Students do | Teacher does |
|---|---|---|---|
| 0:00–0:05 | Hook | [...] | [...] |
| ... | ... | ... | ... |

6. Offer the handoff

Ask the teacher — present three options, not just one:

Two paths from here:

• Post to Canvas now — hand the brief to post-lesson-page. Fills the school's lesson template (targets → to slot, formative → assessment slot, learning plan phases → tasks slot) and creates the draft page.
• Build the materials first — if the lesson needs cards, handouts, hook images, exit-ticket slips, etc., I can draft those before we put anything on Canvas. The page can wait.
• Leave it in chat — you copy the brief and iterate on your own.

Branch on the answer:

• Post to Canvas → invoke the post-lesson-page workflow, passing the brief as established context. post-lesson-page should skip its own Stage-1-style thinking step (everything's done) and go straight to template slot rendering. The competency alignment and module context already exist.

• Build materials first → ask which materials are needed and draft them in chat. Don't volunteer to write files — keep the materials as chat output the teacher can copy. Offer to chain into post-lesson-page after materials are locked.

• Leave it in chat → confirm the brief stays where it is.

If the teacher mentioned specific materials in step 1 ("we'll need case cards," "I want hook images for the opener"), default to offering the materials path first.

Common mistakes to avoid

• Don't draft EUs and EQs from scratch when the lesson belongs to an existing module. Borrow from the module's outcomes page. A lesson advances a module-level EU; it doesn't reinvent one per class.

• Don't over-scale Stage 2. One sharp formative check beats a four-part assessment rubric at the lesson level. If you're tempted to add a summative here, that's plan-assessment's territory.

• Don't write activities (Stage 3) before targets and evidence (Stages 1–2). That's the whole point of backwards design. If the teacher resists and just wants to brainstorm activities, push back gently once — "We'll plan the activities, but we'll get there faster if we know what students should be able to do by the end" — then yield if they insist. Don't be a UbD purist if the teacher needs a quick scaffold.

• Don't pack Stage 3. A 60-minute lesson with 8 phases of 7.5 minutes each is brittle. Leave breathing room.

• Don't produce Canvas HTML in this skill. This is a brief, not a page. post-lesson-page handles template-specific HTML output.

• Don't skip checkpoints. Stage 1 must be confirmed before Stage 2; Stage 2 before Stage 3. UbD silently fails when stages misalign — the checkpoint cadence is the safeguard.

• Don't shoehorn competencies. If the lesson doesn't naturally exercise a competency, don't list it. Saying every lesson hits all 9 makes the framework meaningless.

• Don't fabricate module / course context. If the teacher didn't tell you what module this belongs to, ask — don't infer one and produce a brief that misaligns with a real Canvas unit plan.