PromptFork

Design a fair take-home assignment with a grading rubric and time-box guardrails

Produces a role-specific take-home assignment plus a calibrated grading rubric that respects candidate time and scores real on-the-job signal — an interview, not free work for the company.

Open in Studio
Prompt
You are a senior engineering interviewer who designs take-homes that are fair, time-bound, and signal-rich. A take-home is a candidate's unpaid evening — treat their time as precious.

Role context:
- Role / level: [TITLE, LEVEL — e.g. 'Senior Frontend Engineer']
- The real on-the-job signal that matters: [LIST — e.g. 'API integration, state mgmt, accessibility, code clarity']
- Tech stack the candidate may use: [LANGUAGES / FRAMEWORKS / OR 'THEIR CHOICE']
- Target time box (must be honest): [E.G. '3-4 hours max']
- Format: [BUILD A SMALL APP / EXTEND A REPO / WRITE A DESIGN DOC / DEBUG A PROVIDED REPO]
- What you must NOT request: [E.G. 'no proprietary data', 'no production-scale feature']

Design the assignment with:
1. A one-paragraph brief the candidate reads first: the scenario, the deliverable, the time box, and the 'we are evaluating X, not Y' line so they optimize for the right thing.
2. The concrete task — scoped to the stated time box, with a clear definition of done. If the task cannot fit the time box, cut scope, not the box.
3. What to submit and how (repo, zip, README expectations) and what NOT to spend time on (no need to deploy, no need to gold-plate).
4. A graded rubric: 4-6 criteria, each with 'Below bar / Meets bar / Exceeds bar' described in observable, concrete terms (e.g. 'tests cover the happy path AND one edge' = Meets). No criterion rewards only aesthetics or only volume.
5. A 'we do not penalize' list: time spent, choosing a familiar stack over a fancy one, asking clarifying questions, reasonable assumptions stated in the README.
6. Anti-cheat and ethics notes: how you confirm it is their work without treating them as guilty (a short walkthrough call, not surveillance), and a reminder that the assignment must not be real production work the company will use.

Fairness rules:
- The time box is a ceiling, not a floor. A candidate who hits 'Meets bar' in 2 hours is stronger than one who grinds 8.
- Do not request anything that gives the company shippable product for free. The deliverable is an interview, not unpaid labor.
- Make the task role-realistic and free of hidden traps that only insiders would know.
- Flag any part of the task with disparate-impact risk and propose a fairer alternative.

Output: candidate-facing brief, the task spec, submission guide, the graded rubric, the do-not-penalize list, ethics notes.

Success signal: the output is good only if the task honestly fits the stated time box, the rubric is observable at every bar level, and the deliverable is explicitly not free work for the company.

Use case

Use when a take-home is the right signal for a role and you want it fair, scoped to a real time box, and gradable across candidates.

When to use this

When designing a hiring loop for a role where a small build predicts fit better than live coding. Not for every role — weigh candidate burden first.

Follow-up prompts

  • Convert the rubric into a structured scorecard graders fill in independently.
  • Write the candidate-facing email that sets the time box and expectations fairly.
  • Design a 30-minute walkthrough-call script to verify authorship without interrogation.
#recruitment#interviewing#hiring#take-home#assessment
Source
promptfork seed
License
CC-BY-4.0
Published
6/22/2026

More prompts you might like

Build a calibrated technical interview question bank for a role

Generates a structured, role-specific interview question bank with signal-rubric scoring and bias guardrails — for interviewers, not candidates cramming.

#recruitment#interviewing
New

Aggregate interviewer debriefs into a fair, calibrated hiring decision

Turns scattered post-interview debriefs from a panel into one structured, bias-guarded score aggregation that surfaces agreement and disagreement honestly and makes the hire/no-hire call defensible.

#recruitment#interviewing
New
Editor’s pickJournaling & Self-ReflectionSeed

Turn ChatGPT into a gentle shadow work guide

A prompt that makes AI lead a real shadow-work session — one probing question at a time, reflecting patterns back, ending with an integration practice.

0002

Prompt: an endless creative writing prompt generator

Give it your genre and tone; get original story sparks on demand — each with a character, a situation, and built-in conflict.

0001

Midjourney logo & brand mark — scalable marks that pass the favicon test

Vector-ready logos built on negative space and geometric precision — includes the favicon scalability test, two style examples (monoline vs emblem), and the Ideogram workflow for adding text that doesn't look garbled.

0001
Editor’s pickGame DevelopmentSeed

Open-world (GTA-style) game build prompt

Scopes a 3D open-world prototype realistically — character controller + drivable vehicle + map first, bigger systems phased.

0001