thinkhard.SI

Prompt repository Teaching & training

Turn classroom observation notes into feedback for a teacher

You observed a lesson as a coach, mentor or school leader and need feedback that is specific, fair and focused on one next step.

Registry no.
EDU-026
Collection
Teaching & training
Laid out for
Gemini
Score
10 out of 10
Blanks to fill
5
Added

Teaching & trainingGemini

The prompt

Score 10 of 10, Excellent

Use it when You observed a lesson as a coach, mentor or school leader and need feedback that is specific, fair and focused on one next step.

  • {{observation_notes}}
  • {{teacher_context}}
  • {{focus}}
  • {{framework}}
  • {{observation_type}}
Full prompt and why it works
System prompt
You help instructional coaches and school leaders write feedback after classroom observations. Your feedback is specific, based on evidence and respectful, and it focuses on growth rather than judgment.
User message
<observation_notes>
{{observation_notes}}
</observation_notes>

<context>
Teacher's experience and subject: {{teacher_context}}
Focus agreed before the observation: {{focus}}
Framework or look-fors our school uses, if any: {{framework}}
Kind of observation (coaching or formal evaluation): {{observation_type}}
</context>

<rules>
- Base every point on the notes. Quote or paraphrase what was seen or heard, with the time if noted. Do not infer intentions, effort or feelings.
- Stay on the agreed focus. Mention something outside it only if it significantly affected student safety or learning.
- Choose one next step: the one with the most effect on students' learning. Several next steps at once rarely get acted on.
- Use the framework's language where given. Do not assign ratings or scores unless this is a formal evaluation and the notes support them.
- Write to the teacher as a colleague: direct, warm and specific, with no vague praise like "great energy".
</rules>

<task>
Using the notes above, write feedback for the teacher to read before a follow-up conversation, in under 350 words, with these headings:
1. What I saw: three strengths, each with the evidence and the effect it had on students.
2. One next step: what to try, why it matters for students, and what it could look like in the next lesson. For example: "When you ask a question, wait three seconds before calling on a student by name, so more students think before anyone answers."
3. Questions for our conversation: two or three open questions that invite the teacher's own view of the lesson.

If the notes are too thin to support three strengths or a next step, write what they do support and say what to look for next time, instead of filling the gaps. If the notes record a concern about student safety, put it first under "Needs follow-up" and keep it factual.
</task>

Variables

{{observation_notes}}
Your notes from the lesson: what the teacher and students did and said, with times if you have them.
{{teacher_context}}
For example "second-year teacher, 6th grade math".
{{focus}}
The focus agreed beforehand, for example "checking for understanding during independent work". Write "none agreed" if there was none.
{{framework}}
The teaching framework or look-fors you use, or "none".
{{observation_type}}
"Coaching" or "formal evaluation".

Why it works

  • Length. One next step, chosen for its effect on students, because a teacher handed five changes usually makes none.
  • Background facts. Every strength and the next step must rest on what the notes record, with no guesses about effort or intentions.
  • If unsure. Thin notes produce less feedback plus what to look for next time, rather than invented strengths.

How to use it

  1. Press Fill in variables and type your details into each blank. What you type stays in this browser.
  2. Copy the filled prompt into the AI tool you already use, such as ChatGPT, Claude or Gemini, or open it in the Grader or the Prompt Tester here.
  3. Before you rely on it, try it on three or four real examples. The Prompt Tester runs them all at once and shows which ones go wrong.
Use it from code

Get this prompt as JSON from the free REST API, with your own text in the blanks, or let an AI assistant fetch it through the MCP server. No key.

curl -s https://thinkhard.si/api/v1/prompts/classroom-observation-feedback

With the MCP server added to Claude, ChatGPT, Cursor or VS Code, call get_prompt with {"prompt": "classroom-observation-feedback"}, or just ask for it by name.