AI agent for secondary school teachers
Essay Feedback Rubric Consistency Agent
Consistent rubric scores and specific feedback for every essay in a set
What it does
Teachers grading a tall stack of essays get tired, and the same essay can earn a different score at 9 pm than at 4 pm. This agent pulls the essays and rubric from the learning system and drafts feedback for each rubric row, quoting the student's own lines as evidence. The teacher scores each essay using that draft. Then the agent compares scores across the whole set, grouping essays with similar evidence. Any score more than one level away from similar papers is listed with side-by-side quotes so the teacher can re-read it, and the agent re-runs the comparison after changes. The teacher always sets the final grade and releases it. Edge case: if an essay looks heavily copied or AI-written, the agent does not accuse anyone; it lists the passages that need a closer look and leaves the decision to the teacher.
How it works
Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.
Read the steps as a list
- Essay assignment closes
- Pull essays and the rubric from the LMS
- Draft feedback per rubric row with quoted evidence
- Teacher scores each essay using the draft feedbackThe agent waits here for your OK.
- Compare scores across essays with similar evidence
- Is every score within one level of similar essays?If not: list outliers with side-by-side quotes for the teacher to re-read. Back to step 4.
- Teacher releases grades and feedbackThe agent waits here for your OK.
- Graded set with consistent feedback
How it decides
It compares each essay's rubric scores to essays with similar evidence. A paper is flagged when its score differs by a full level from the closest matches.
- Feedback must quote the student's own words for each rubric row
- Score gaps of one full level from similar essays are flagged
- Possible integrity issues are listed, never labeled
Make it yours
Every agent is a starting point. You choose these settings for your own situation.
- Rubric to use
- Outlier gap that triggers a flag (default 1 level)
- Feedback length per row (default 2 sentences)
- LMS used (Google Classroom, Canvas, Schoology)
What keeps you in control
It always asks you first
- Teacher sets every grade
- Teacher releases grades to students
Hard limits
- Never posts grades or feedback to students on its own
- Never labels a student as cheating
It stops when
- Done: all essays scored with no unresolved outliers
- Stop: rubric is missing
Set it up
We guide you through the set-up, step by step
Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.
- One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
- The agent then walks you through connecting your own data, one source at a time
- A downloadable copy with the flow chart, the rules and the full guide