Reference

Prompt bank

Every prompt follows the same frame: context, task, constraints, evidence, output. Replace the bracketed parts with your own details.

Before you paste

Remove learner names and any confidential information. Every output still passes the quality gate: accuracy, alignment, fairness, privacy, transparency.

01 · Plan and explain

Lesson plan from an outcome
Context: I teach [module] to [learner group] in a TVET setting, [session length] practical session.
Task: Plan the session so learners can [observable outcome].
Constraints: [equipment/space limits]; mixed prior experience; plain language.
Evidence: State the check that shows each learner met the outcome.
Output: A timed plan with educator moves, learner actions and checks. Mark where my professional judgement is required.

02 · Ground in your own sources

Source-bound study guide
Context: The attached files are the only permitted sources for [module], [learner group].
Task: Build a study guide for [topic] using only those sources.
Constraints: Do not add outside facts; if a source is silent, say so explicitly.
Evidence: Cite the source and section for each key point.
Output: Key points, 8 practice questions with answers, and a list of gaps I should teach directly.

03 · Differentiate

Three access levels, one core task
Context: [module], [learner group]; some learners read below level and some are ahead.
Task: Take the task below and produce three versions that keep the same assessed outcome.
Constraints: Same criteria for all versions; no reduced expectations, only different support.
Evidence: For each version, state the scaffold added or removed and why.
Output: Supported / core / stretch versions plus a one-line note on when to use each.

Task:
[paste your task]

04 · Create practice

Scenario-based practice items
Context: [module], [learner group], TVET workshop context.
Task: Write [6] scenario-based questions on [topic], each testing one named decision learners must make on the job.
Constraints: Realistic local scenarios; no trick questions; one correct answer per item.
Evidence: For each item give the answer, the rationale and the common misconception it exposes.
Output: A numbered list, then a short table of misconceptions to address in teaching.

05 · Draft feedback

Criterion-linked feedback draft
Context: [module], practical assessment, criteria listed below. The learner's work is summarised below.
Task: Draft feedback the learner can act on before the next attempt.
Constraints: Maximum 120 words; specific and respectful; no marks or grades; no invented observations.
Evidence: Link every comment to one named criterion.
Output: Two strengths, one priority to change, and the next single action.

Criteria: [paste]
Observed: [paste your notes]

06 · Analyse waste in a process

Waste analysis of a teaching workflow
Context: The steps below are how [teaching process] actually runs, from [trigger] to [learner outcome].
Task: Identify where learner value is created and where waste occurs.
Constraints: Use the eight wastes — defects, overproduction, waiting, unused talent, transport, inventory, motion, overprocessing. Do not propose new tools.
Evidence: For each waste, state the likely cost in time, quality or learner effort.
Output: A table of Step | Value or waste | Type | Cost, then the single constraint to attack first.

Steps:
[paste your steps]

07 · Design a future state

Future-state options
Context: Current process and constraint below, for [learner group].
Task: Propose three future-state options that reduce lead time without lowering standards.
Constraints: Keep educator judgement on assessment decisions; no new paid tools; describe changes in flow, not features.
Evidence: For each option state the expected effect on lead time, quality and workload, plus the main risk.
Output: Three options ranked by effort, and one two-week pilot for the top choice.

Current process and constraint:
[paste]

08 · Standard work and instructions

One-page standard work
Context: [recurring teaching task] in [module], done by me and occasionally by a relief instructor.
Task: Write one-page standard work for it.
Constraints: Maximum 10 steps; plain language; no tool-specific jargon.
Evidence: For each step, the quality criterion and the visible check.
Output: A table with Step | What good looks like | Judgement needed | Common failure. Flag anything to verify against local policy.

09 · Assessment: item set with a thinking-level spread

20 items across Bloom levels
Context: [module], [learner group], TVET; assessment covers [topic].
Task: Build 20 assessment items with this spread: 20% recall, 25% understanding, 25% application, 15% analysis, 10% evaluation, 5% creation.
Constraints: Workshop-realistic contexts; one correct answer per objective item; no trick questions.
Evidence: For each item, name the level and the outcome or standard it measures.
Output: A table — No | Item | Level | Outcome | Answer | Marks. After the table, list the three hardest items and why.

10 · Assessment: distractor analysis

Make every wrong option diagnostic
Context: The objective items below are for [module], [learner group].
Task: Rewrite the options so each distractor represents ONE specific learner misconception.
Constraints: Keep the stem and the correct answer unchanged; no "all of the above"; plausible distractors only.
Evidence: After each item, state in one sentence what a learner choosing each distractor likely misunderstands.
Output: Revised items plus a short misconception table I can teach from.

Items:
[paste]

11 · Assessment: analytic rubric

Four-level observable rubric
Context: [practical task] in [module], assessed by me and occasionally a relief instructor.
Task: Build a four-level analytic rubric.
Constraints: Maximum 4 criteria; each descriptor must describe OBSERVABLE behaviour, not abstract traits; avoid "good" or "satisfactory" without explanation; maximum 25 words per descriptor.
Evidence: State the evidence an assessor would see or hear for each level.
Output: The rubric table plus a mark allocation totalling 100, and a note on where my judgement is required.

12 · Assessment: let AI interview you first

Questions before generation
Context: I want to build a formative assessment for [learner group] in [module].
Task: Before generating anything, ask me 5 questions to understand my group, the topic and the learners' current level.
Constraints: Ask one question at a time; wait for my answers before producing anything.
Evidence: Summarise what you learned before you generate.
Output: After my answers, the assessment plus the check that shows the outcome was met.

13 · Designing activities: one hour of active practice

Activity sequence from an outcome
Context: [module], [learner group], [session length] in a workshop/lab with [equipment and space limits].
Task: Design an activity sequence where learners spend most of the time doing, not listening, so they can [observable outcome].
Constraints: Maximum 10 minutes of instructor talk; groups of [n]; equipment sharing must be realistic; include a safety brief.
Evidence: For each activity, state what the instructor observes to know it is working.
Output: A timed sequence — minute range | learner action | instructor move | check — plus one fallback if equipment fails.

14 · Designing activities: three access levels, one outcome

Supported / core / stretch activity
Context: [module], [learner group]; mixed prior experience and reading levels.
Task: Turn the activity below into supported, core and stretch versions.
Constraints: Same assessed outcome and same criteria for all three; change support, not expectations; keep the same step numbering so I can supervise all three at once.
Evidence: For each version, state the scaffold added or removed and why.
Output: Three versions plus a one-line note on when to use each.

Activity:
[paste]

15 · Designing activities: interactive game brief (Canvas)

One-page learning game
Context: [module], [learner group]; used on a classroom projector and on learner phones.
Task: Use Canvas to build a one-page [matching / quiz / sequence-ordering] game on [topic].
Constraints: [8] items; plain English; large text; works offline once loaded; no learner data collected.
Evidence: Show the correct answer and a one-line explanation after each wrong attempt.
Output: A working game with scoreboard, timer and replay button, plus a printable answer key for me.

16 · Materials: learner-facing handout

One-page handout from your own sources
Context: The attached sources are the only permitted content for [module], [learner group].
Task: Write a one-page learner handout for [topic].
Constraints: Plain English at [reading level]; maximum [400] words; every technical term defined on first use; printable in black and white.
Evidence: Cite the source section for each key point and flag anything absent from the sources.
Output: Title, 5 key points, one worked example, a labelled diagram description, and 3 self-check questions with answers.

17 · Materials: worksheet from an infographic prompt

Notes, table and practice on one sheet
Context: [module], [learner group]; sources uploaded to my notebook.
Task: Produce one worksheet sheet combining condensed notes, a comparison table and objective practice on [topic].
Constraints: One page; space for learners to write; icons and soft colours for supported learners, dense layout for revision; no outside facts.
Evidence: Mark which source section each section of the sheet comes from.
Output: The worksheet layout plus a separate answer key and a note on what I must verify against local policy.

18 · Materials: slide deck first draft

Presenter-ready slides
Context: [module], [learner group], [session length]; slides shown on a workshop projector.
Task: Draft a slide deck for [topic].
Constraints: One main idea per slide; maximum 25 words of slide text; every slide carries a workshop example; plain English.
Evidence: For each slide, add a speaker note with the check or question I ask the group.
Output: Slide title, slide text and speaker note per slide, plus 3 slides marked as optional if time runs short.

19 · Materials: accessible alternatives

Equivalent routes to the same learning
Context: [module], [learner group]; some learners read below level, some missed the session, some need audio.
Task: Produce equivalent alternatives to the material below without lowering the standard.
Constraints: Keep the same outcome and terminology; no alternative may become the only route; include alt text and caption guidance.
Evidence: For each alternative, state which barrier it removes.
Output: A lower reading-level version, an audio/script version, a visual summary description, and a printable version.

Material:
[paste]
Finished this section?

Mark it so your progress is kept in this browser.

Lean Digital Instruction — a practical workshop for TVET educators. 12-13 August 2026, Gamification Centre UNIMAS.
Facilitated by Dr Chuah Kee Man · www.chuahkeeman.com