Category · Coding & Development

Prompts for Coding & Development

Refactors, test suites, code review, migrations and debugging with the context an LLM needs.

Popular: Blog outline Code review Midjourney Cover letter SQL

56 Prompts

14 Categories

12 AI models

16 Featured

AI Directory Companies Comparisons Alternatives Best-of Guides Use Cases Collections AI Tools AI Models AI Courses

4 prompts

Updating…

Featured

Refactor With the Constraints Stated

Coding & Development

A refactor that respects your language version and does not silently change behaviour.

ChatGPT Claude GitHub Copilot

Code Intermediate 3 380

Open

Refactor the [LANGUAGE] code below.

CONSTRAINTS — these are hard:
- Language/runtime version: [VERSION]. Do not use anything newer.
- Behaviour must be identical, including error cases and edge cases.
- Keep the public signature unless I say otherwise.
- Match the surrounding style: naming, comment density, idiom.

Goals, in priority order: correctness, then readability, then performance.

Return:
1. The refactored code
2. A numbered list of every change and the reason
3. Anything you noticed that is a real bug — separately, NOT fixed inline, so I can decide
4. Any behaviour you could not preserve, stated plainly

If the code is already fine, say so instead of churning it.

CODE:
[CODE]

Featured

Code Review That Ranks by Severity

Coding & Development

A review that separates real bugs from taste, so you know what must be fixed.

ChatGPT Claude

Code Advanced 2 400

Open

Review this [LANGUAGE] diff as a senior engineer on the team.

Sort every finding into exactly one bucket:
- BUG: it will produce a wrong result or crash. Give the concrete input that triggers it and the wrong output.
- RISK: it is correct today and fragile tomorrow. Say what change would break it.
- SIMPLIFY: same behaviour, materially less code or less indirection.
- TASTE: you would write it differently. Keep this section short.

Rules:
- No finding without a failure scenario or a concrete replacement.
- Do not report style the file already violates consistently — match the surrounding code.
- If you find nothing in a bucket, write "none" rather than manufacturing one.

DIFF:
[DIFF]

Tests for the Cases That Actually Break

Coding & Development

Generates a test suite weighted toward boundaries and failure paths, not happy paths.

ChatGPT Claude GitHub Copilot

Code Intermediate 2 360

Open

Write [FRAMEWORK] tests for the code below.

Weight the suite toward what breaks in production:
- Boundaries: empty, single element, maximum, off-by-one
- Nulls and missing keys at every level the code touches
- Wrong types where the language allows them through
- Concurrent or repeated calls, if state is involved
- The error paths — assert the message, not just that it threw

One happy-path test is enough.

For each test: a name that states the expectation, arrange/act/assert clearly separated, and no shared mutable state between tests.

After the suite, list what you could NOT test without changing the code, and what the smallest change would be.

CODE:
[CODE]

Browse Prompt Categories

Every prompt in the library, grouped by the job it does — pick the shelf, then the prompt.

Writing & Content

Blog posts, newsletters, scripts and long-form drafts you can edit rather than start from scratch.

4 prompts

Marketing & Ads

Ad copy, landing pages, launch emails and positioning built around a real offer and audience.

4 prompts

SEO & Search

Keyword clustering, briefs, metadata and on-page rewrites aimed at search intent.

4 prompts

Coding & Development

Refactors, test suites, code review, migrations and debugging with the context an LLM needs.

4 prompts

Data & Analysis

SQL, spreadsheet formulas, cohort analysis and turning a dataset into a readable finding.

4 prompts

Design & Image

Image prompts with subject, lens, lighting and style separated so you can swap one at a time.

4 prompts

Video & Motion

Shot lists, scene descriptions and camera language for text-to-video models.

4 prompts

Business & Strategy

Positioning, pricing, competitor teardowns, board updates and decision memos.

4 prompts

Education & Learning

Lesson plans, explanations at a chosen level, quizzes and study schedules.

4 prompts

Productivity

Meeting notes, inbox triage, weekly planning and turning a mess into a next action.

4 prompts

Customer Support

Reply templates, macro libraries, escalation summaries and tone-matched apologies.

4 prompts

Social Media

Hooks, threads, captions and repurposing one idea across every platform.

4 prompts

Research & Insights

Literature summaries, interview guides, synthesis and steel-manning the other side.

4 prompts

Career & Resume

Resume bullets, cover letters, interview prep and performance-review self-assessments.

4 prompts

See all categories

Questions

Prompt FAQs

Still stuck? Our team answers every message.

Still have a question? We usually reply within one working day.

How do I use these prompts?

Copy any prompt with one click, paste it into your AI assistant, then replace the bracketed blanks — [TOPIC], [AUDIENCE] and so on — with your own details before sending. The blanks are listed on every prompt page so you know what to prepare.

Which AI models do these work with?

Each prompt lists the models it was written for. The text prompts are largely model-agnostic and run well on ChatGPT, Claude, Gemini and other assistants; the image and video prompts use syntax specific to their tool, such as aspect-ratio flags, so those are worth matching to the model named on the card.

Are the prompts free to use?

Yes. Every prompt in the library is free to copy, edit and use commercially. You do not need an account to copy one.

Why do some prompts tell the model to say “I do not know”?

Because a confident wrong answer is more expensive than an admission of uncertainty. Several prompts here ask the model to mark unverified claims or to stop and ask rather than guess — that instruction is doing real work and is worth keeping when you edit them.

Still have a question? We usually reply within one working day.

Prompts for Coding & Development

Refactor With the Constraints Stated

Code Review That Ranks by Severity

Tests for the Cases That Actually Break

Debug From a Stack Trace

Browse Prompt Categories

Writing & Content

Marketing & Ads

SEO & Search

Coding & Development

Data & Analysis

Design & Image

Video & Motion

Business & Strategy

Education & Learning

Productivity

Customer Support

Social Media

Research & Insights

Career & Resume

Prompt FAQs

How do I use these prompts?

Which AI models do these work with?

Are the prompts free to use?

Why do some prompts tell the model to say “I do not know”?

Popular Searches

Try Something New

Trending Tools