Prompt Workshop
Why it is written this way
Tap a block to see its paragraph
You are an experienced teacher designing a performance task for {{academic subject}}. Your priority is to create an objective assessment where every score can be clearly justified to students and parents based on concrete evidence.
I want to design an assessment for {{student grade level}} students on the topic "{{assignment topic}}". It must be completed within 2 class periods plus one individual submission (individual work only, not group work).
Do not generate everything at once; proceed in three distinct steps. Step 1 is Task Design, Step 2 is the Grading Rubric, and Step 3 is the Student Assignment Handout. Stop after completing each step, and only proceed to the next step when I reply with "Next".
In Step 1, provide two different task options. For each option, clearly specify: the actual student activity, the final deliverable format, the estimated completion time, and which specific part of the standard below it measures. In Step 2, based on the option I select, create a 4-level rubric table with 3 evaluation criteria. In Step 3, write the student handout in this order: Assignment Overview, Due Date & Submission, Rubric Summary, and Common Mistakes to Avoid.
Do not invent or reference standard codes or curriculum document numbers. Base your reasoning strictly on the text provided below, without adding external content. Do not use real student names or identifiable personal scenarios in the task.
Standard: """ {{achievement standard}} """
- Role Who answers ×1
- Material What the AI needs to know ×2
- Work What to do, and how ×1
- Output What comes back, in what shape ×2
When asking AI to write a full assessment plan all at once, the task, rubric, and student handout often fail to align. For example, the prompt might describe a research paper, but the rubric evaluates presentation posture, and the handout explains a different task altogether. Generating all three at once causes the AI to lose track of consistency across documents.
Read the full explanation · 3 paragraphs
This prompt resolves that issue by breaking down the workflow into three sequential steps. You finalize the task first, generate the rubric based on that chosen task, and finally summarize the rubric into student-facing guidelines. Because each output serves as the input for the next stage, the documents stay perfectly aligned. The instruction "only proceed when I reply with 'Next'" establishes a necessary pause.
In the role section, specifying "assessments where scores can be clearly justified" is critical for evaluation design. This constraint prevents the AI from proposing open-ended, subjective creative tasks that are hard to grade objectively, steering it toward tangible, observable student artifacts. In the format section, asking for two task options in Step 1 gives the educator room to choose rather than forcing revisions on a single imperfect idea.
Finally, the constraints prevent two common pitfalls: hallucinated curriculum standard codes (which AI frequently invents) and overly specific, identifiable student scenarios that may leak bias.
Unfamiliar terms? See Aha AI: chain-of-thought, hallucination
Compared with a bad example
Create an 8th grade science performance assessment plan. Topic: reducing cafeteria food waste.
This produces a generic, one-page assessment plan. While it contains a task, rubric, and handout, they often contradict one another—such as rubric criteria measuring skills the task never asked students to demonstrate. It might also hallucinate official-looking standard codes that do not actually exist.
Variations
When you only need a rubric fast
Use this when the assignment prompt is already set and you just need a reliable rubric.
Show prompt
Please generate only the grading rubric for a {{student grade level}} {{academic subject}} assessment on "{{assignment topic}}".
Create a 4-level scoring table with 3 evaluation criteria. In each cell, write one clear sentence describing the observable student behavior in the final deliverable. Under each criterion, include one line explaining which part of the provided standard it assesses.
Standard: """ {{achievement standard}} """
Reviewing an existing task for fairness
Use this when you need a pre-implementation audit. Fairness and grading discrepancies are difficult to fix once the assignment has been launched.
Show prompt
I previously ran a performance task in {{student grade level}} {{academic subject}} on "{{assignment topic}}". Before running it again, please help me audit potential issues.
Identify: 1) areas where students could easily copy/paste from the internet without thinking, 2) factors where access to home resources or materials could create an unfair advantage, and 3) points where different graders might score inconsistently. For each issue identified, provide a one-sentence concrete fix to improve the task design.
Model notes
Because AI models can hallucinate standard codes, paste the actual standard text rather than codes, and add your official curriculum codes directly into the final output.
In multi-step conversational prompts, long message threads can sometimes cause the AI to drift from earlier decisions. When prompting to move to Step 2, re-pasting your selected task option ensures the rubric matches the assignment perfectly.
Last updated 2026-09-02 · Found a mistake? Let us know