Self-grading Knowledge Check
Build an internal tool that delivers an approved quiz, grades it instantly, shows what was missed, handles retries, and keeps proof of completion - no more grading paper or chasing redos.
A logged-in quiz tool that auto-grades instantly, shows per-question feedback, enforces retries, records every attempt, sets pass/fail, and exports a results CSV.
Before you start
- An approved question bank or quiz (questions, answer choices, correct answers)
- Your pass mark and retry rules
- An assignment list (who has to take which quiz)
The problem this kills
You hand out a knowledge check, then you collect it, grade it by hand, and chase the people who didn't pass to "redo it." Scores live in your head or a spreadsheet. When an auditor or your boss asks "can you prove everyone passed the safety quiz?", you start digging through paper or email threads.
It is slow, it is error-prone, and there is no clean record. Worse, nothing stops someone from taking the same quiz five times until they luck into a pass - and you'd never know.
What you'll build
A small web app your team logs into. A trainer (you) loads an approved quiz, sets the pass mark and retry rules, and assigns it to people. Each employee logs in, takes the quiz, and gets graded the instant they submit - with their score and exactly which questions they missed. Retries are handled automatically per your rule, and the quiz locks after the last allowed attempt until you step in. Every attempt is recorded, pass/fail is set automatically, and you can export the whole results table as a CSV anytime.
What's inside the Implementation Plan
The plan opens by interviewing you about your business - your current quiz process, how your people and quizzes are named, your real pass marks and retry limits, your peak volumes, and your awkward edge cases. It uses your answers to shape the data model and rules, so you get a tool tailored to how you actually run training, not a generic quiz template you have to bend to fit.
From there it walks you, prompt by prompt, through building the whole thing: login, the trainer screens for loading and approving a quiz, the employee quiz-taking screen with randomized answer order and instant per-question feedback, retry handling and quiz lock-out, the complete attempt log, and the CSV export. Each step ends with a ready-to-paste prompt for your AI coding agent.
The governance it includes (this is the point)
- Login so only your team can use the tool.
- Row-level security so people only ever see their own organization's data and their own attempts.
- A full audit trail - who took what, when, what they scored, and any change you make.
- A human-in-the-loop approval gate: the trainer approves the quiz and the pass/retry rules before anyone is assigned, and any score override or extra-attempt grant requires your approval and is logged.
- Duplicate guards so the same attempt can't be recorded twice (keyed on employee + quiz + attempt number).
Who it's for
Ops trainers and team leads who today grade paper quizzes or nag people to redo them - and who need clean, exportable proof that everyone actually passed.
You've got this - paste the first prompt and let the agent interview you.