Hire an AI student rubric feedback writer

Everything it can hand back

Your task sets the actual outputs. The specialist picks the methods that cover them.

• A rubric in plain words for each level, for you to approve before marking
• A level for each criterion and feedback for every piece
• The two or three edge cases, with the case for each level
• Class patterns with two quoted examples each, pointing to the next lesson
• Students whose level rose or fell since last time
• Questions about work that looks copied, never a verdict

How it does the job well

• Base people decisions on job-related evidence, applied the same way to everyone.
• Write in plain, respectful language the person affected could read.
• Keep personal data to what the task needs, and handle it privately.
• Check local employment law and company policy before advising on hiring, pay or exits.

Mistakes it avoids

• Writing the fix into the note.
• Changing the rubric halfway through a batch.
• Marking the thirtieth piece faster than the first.
• Treating a quiet improvement as nothing to say.
• Biased wording in job ads, reviews or interview questions.
• Plans with no owner, timing or way to check progress.

What a good result looks like

An example, not your task: You teach a class or run a course and thirty pieces are waiting to be marked.

• The batch. One line with pieces marked, levels reached, edge cases and patterns found.
• Per piece. A level for each criterion, the gaps, and feedback that names the problem.
• Edge cases. Pieces between two levels, with the case for each level, left for you to decide.
• Class patterns. Each criterion a third or more of the class missed, with two quotes and the next lesson.
• Growth. Students whose level rose or fell since their last piece.
• Questions for the teacher. Work that looks copied or out of character, written as a question, never a verdict.

How the work is checked

• Every piece was marked against the same saved rubric.
• Every note names a problem; none tells the student what to write.
• Edge cases were left to the teacher.
• Patterns are counted from the marks, with quoted examples.
• States which checks passed, failed or could not run, and names missing inputs.

These checks have not been run against a model yet. Your AI uses them to check its own work.

A second opinion

Before it hands back, your AI reviews the draft the way Ed Catmull and Jane Austen would. This is a way of reviewing the work. It does not mean they wrote or endorse it.

Capabilities

• Build the rubric once per task. Clear wording at each level, written for a student to read, approved by you before any marking.
• Mark each piece. A level per criterion and the gaps, with pieces on the edge between two levels marked for you.
• Read the writing. For essays and stories, names the weak lines and asks whether the student is fine with them.
• Sharpen every note. Each note names a problem, offers any fix only as an example, and cuts anything generic.
• Find the patterns. Any criterion a third or more missed becomes a class pattern with two quotes and the lesson it points to.
• Log and hand back. Saves each piece and the student rows, then stops. You send.

How the role works

1. Checks your task, context and available tools.
2. Picks the methods needed and prepares a work plan.
3. Carries out supported tasks, with each method's limits and review gates.
4. Checks the real deliverables and corrects supported faults.
5. Hands back files, evidence and unresolved gaps.

Your specialist's working setup

Download the base specialist without a task form or account. README.md and SETUP.md explain how to use the folder. SOUL.md defines its scope. INTAKE.md asks missing questions inside your AI. PLAYBOOK.md, SKILLS.md, BEST_PRACTICES.md, CHECKS.md and HANDOFF.md guide the work and its checked deliverables.

Complete associated skills, supporting files and licences are included. Choose only the methods needed for the actual task. Keep context, working files and accepted feedback in your AI project.

Use this specialist through MCP or the API

Call prepare_specialist with role_id student-work-marking-desk only. No task is required to load it. Your AI asks what job you need done, what you want back and which relevant context it can use.

For task-led hiring, call find_specialist, explain the suggested specialists and let the user accept or change the team.

Complete ZIP: Download base specialist. API: GET /api/v1/specialists/student-work-marking-desk/pack. The host AI supplies its own model, tools and execution; a pack does not connect accounts or run a task.

Inputs and helpful context

The specialist starts from the task brief. Only active methods require their inputs.

For the primary method, when used

• The task and what a good piece shows the student learned
• Each piece of student work, as text or a file, with a student id
• Your grade scale and any comment wording your school needs (optional)
• Your saved rubric from an earlier batch (optional)
• Past batches of the same class (optional)

Helpful context (optional)

• Your school or programme's required comment wording
• Exemplars at each level from a past batch, so the rubric has anchors

Tools and access

• Claude Code: It can keep one folder per task, one file per piece and the student rows between batches.
• Any good chat AI: Paste the agent, the task and a few pieces at a time. Keep the rubric in a note.
• Any good chat AI: Paste the table into a document or sheet when it is done.
• Any good chat AI: Planning, questioning and structuring are text work any strong chat AI does well with this skill.
• Any good chat AI: Sorting and rewriting comments is text work any strong chat AI does well with this skill.
• A shared document: Reviewers can add their answers to the questions beside each note.

Functions and deliverables

• Student rubric feedback writer: Review student work against one approved rubric. Get evidence-linked feedback, class patterns and borderline cases for the teacher.
  • A rubric in plain words for each level, for you to approve before marking
  • A level for each criterion and feedback for every piece
  • The two or three edge cases, with the case for each level
  • Class patterns with two quoted examples each, pointing to the next lesson
  • Students whose level rose or fell since last time
  • Questions about work that looks copied, never a verdict
• Task marking guide writer: Describe the task and learning goal. Get a marking table with clear criteria, four levels and feedback lines.
  • A rubric table
  • A feedback line for each level
  • A tip on how to use it
• Student marking and feedback reviewer: Share the learning goal, student work and feedback draft. Get a marking guide, evidence-linked gaps, possible causes and clearer feedback.
  • A single-point rubric with open columns for evidence
  • The exact gaps between the work and the rubric
  • Likely causes of the mistakes, each with a feedback or teaching fix
  • A check of your comments with a better wording for the weak ones
• Long-form draft reviewer: Share a full draft, its purpose and what matters to you. Get specific notes on strong lines, weak spots and choices to reconsider.
  • A short note on what is working
  • Notes on specific weak lines, each as a question
  • Where the reader leans in and where they drift
  • A leave-alone list and a check of each action against the character
• Draft feedback editor: Turn vague or bossy feedback into clear notes on the problem. Paste your raw comments and get each one rewritten with its place and any fix as an example.
  • Each comment rewritten as a problem note with its place and what the reader lost
  • Fixes kept on their own line, marked as examples only
  • Repeats merged and counted
  • A list of questions to send back to reviewers, and a list of comments that were about you, not the work
• me.md builder: Answer a few questions, or let your AI use what it remembers. Get one private file that tells any AI who you are.
  • One me.md file with up to ten dated sections
  • A list of facts taken from memory, for you to check
  • A short note of what it removed for privacy
  • The two next steps, to save it and set it up
• Ed Catmull: Plan feedback or protect rough creative work using Ed Catmull’s recorded approach. Name the problem, keep the creator’s choice and find hidden issues.
  • Your problem worked on the way Ed Catmull would: their questions, their method, their push-back
  • The answer or the work, in Ed Catmull's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step
• Jane Austen: Edit a scene or character through Jane Austen’s recorded approach. Check motives, sharpen each voice and let readers see what the character misses.
  • Your problem worked on the way Jane Austen would: their questions, their method, their push-back
  • The answer or the work, in Jane Austen's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step

Tasks you can use it for

• You teach a class or run a course and thirty pieces are waiting to be marked.
• Your feedback has started writing the fix for the student.
• You want to know what a third of the class got wrong before you plan the next lesson.
• You mark the same task again and want to see who has grown.

Limits to check before starting

• You need a grade sent or saved in your school system: do that yourself, because this agent never sends or records a grade.
• You mark code, maths problem sets or multiple-choice exams with one right answer: use an answer key or your platform's auto-marking.
• You only have one or two pieces and no rubric: read them yourself and write the notes by hand.

Best way to use this role

Give it one clear outcome, the actual files and a definition of done. Keep the work in one AI project so it can reuse checked context and earlier files.

• Open a chat with the agent. In Claude Code, add it as a skill or save it in .claude/agents. In ChatGPT or Gemini, paste it in.
• Describe the task and what a good piece shows. Before you paste any student work into an AI tool, check your school's data-protection and consent rules. Remove names first and use student ids.
• On the first batch, read the rubric closely and approve it. It is kept for later batches of this task.
• Attach the batch. In a chat AI, paste a few pieces at a time. Claude Code can read a folder of files.
• Settle the edge cases, then read the class patterns before the pieces.
• Approve the feedback and send grades and notes yourself through your school's system.

Read it as data

Read the complete role setup as JSON. Connect your AI to the AI Blocks MCP library and read aiblocks://roles/student-work-marking-desk for its role instructions and pinned sources.

This is a prepared AI role. No task has run. Your AI carries out authorized work using its available tools and reports the actual deliverables and checks.