devprompts logodevprompts

80+ production-tested AI prompts

Stop wasting tokens.
Start shipping faster.

A curated prompt library for full-stack engineers — covering code review, debugging, refactoring, docs, and architecture. Built for Claude, ChatGPT, and GitHub Copilot.

Join engineers who've already saved hours this week

You're already paying for AI.
You're not getting 80% of the value.

Vague prompts, vague answers

Generic inputs return generic output. You re-run the same prompt five times hoping it clicks.

Blank page every time

No system, no history. Every debugging session starts from scratch.

Teammates don't use AI at all

No shared baseline. Half your team is still doing things the slow way.

A prompt system built the way engineers actually work

Not a collection of tips. An opinionated, structured library organized by task — with context-setting headers, chain templates, and variable slots so you can paste and go.

code-review-master.md

Everything in the kit

Code Review
10 prompts
PR Descriptions
8 prompts
Debugging & Diagnosis
12 prompts
Refactoring
10 prompts
Test Generation
8 prompts
API DesignPro
8 prompts
Architecture ReviewPro
8 prompts
Documentation & READMEPro
8 prompts
Interview Prep — BonusPro
10 prompts
FeatureStarterPro
Prompt count4882+
Categories59
Chain templates
Notion template
Architecture prompts
Interview Prep (10 prompts)
FormatPDFPDF + Notion

See what you're getting

Code Review

Master Code Review

You are a senior software engineer performing a thorough code review.

Context:
- Language/Framework: [LANGUAGE_FRAMEWORK]
- PR purpose: [PR_PURPOSE]
- Team experience level: [TEAM_LEVEL]

Review the following code and return findings grouped by severity:
🔴 CRITICAL — must fix before merge (bugs, security, data loss)
🟡 MAJOR — should fix (performance, maintainability)
🟢 MINOR — nice to fix (style, naming, minor improvements)

For each finding provide:
1. Location (file:line if possible)
2. Issue description
3. Why it matters
4. Suggested fix (with code snippet if applicable)

Code to review:
```[LANGUAGE]
[CODE]
```

End with a one-paragraph overall assessment and a MERGE / REQUEST CHANGES / APPROVE WITH NITS verdict.
Debugging & Diagnosis

Stack Trace Interpreter

You are an expert debugger. Interpret the following stack trace and explain it in plain English.

Language/Runtime: [LANGUAGE_RUNTIME] (e.g., "Node.js 20", "Python 3.11", "JVM 17")
Application context: [BRIEF_APP_DESCRIPTION]

Stack trace:
```
[PASTE_STACK_TRACE]
```

Additional context (logs, recent changes, reproduction steps):
[ADDITIONAL_CONTEXT]

Provide:
1. **Root cause** (in one sentence, plain English — no jargon)
2. **Call chain walkthrough**: Step through each frame explaining what was happening
3. **The exact line that failed** and why
4. **Most likely causes** (ranked 1–3 with probability estimate)
5. **Immediate fix**: What to change to resolve the most likely cause
6. **Preventive measure**: How to prevent this class of error in the future
Refactoring

Make This Testable

You are refactoring code to make it unit testable without changing its external behavior.

Language: [LANGUAGE]
Code:
```[LANGUAGE]
[CODE]
```

Identify what makes this code hard to test:
1. **Hard dependencies** — direct instantiation of external services, databases, APIs
2. **Static method calls** — global state access that can't be replaced in tests
3. **Side effects baked into constructors**
4. **Non-deterministic behavior** — random numbers, current time, environment reads
5. **No seams** — nowhere to inject test doubles

For each testability problem:
- Show the problematic pattern
- Apply the appropriate fix:
  - Constructor/parameter injection for dependencies
  - Strategy pattern for variable behavior
  - Clock/random abstraction for non-determinism
  - Pure function extraction for logic buried in I/O

Show the refactored code, then show a sample unit test that would now be possible.

Pick your tier

Starter Kit

$27one-time

The essential prompt library for engineers who want structured, repeatable AI workflows.

  • 48 production-tested prompts
  • 5 core categories
  • Code Review (10 prompts)
  • Debugging & Diagnosis (12 prompts)
  • Refactoring, PR Descriptions, Tests
  • Chain templates included
  • PDF format, instant download
  • API Design prompts
  • Architecture Review prompts
  • Interview Prep bonus
  • Notion template
Most popular

Pro Kit

$67one-time

The complete system — every prompt category, chain templates, Notion integration, and interview prep.

  • 82+ production-tested prompts
  • 9 categories (all of Starter +)
  • API Design (8 prompts)
  • Architecture Review (8 prompts)
  • Documentation & README (8 prompts)
  • Interview Prep — Bonus (10 prompts)
  • Notion template included
  • PDF format, instant download

One-time payment. Instant delivery. No subscription.

Common questions

What AI tools does this work with?

All of them. The prompts are plain text — they work with Claude, ChatGPT, GitHub Copilot Chat, Gemini, and any other AI assistant that accepts text input. No special integration required.

Is this a subscription?

No. This is a one-time purchase. You pay once and own the kit.

What format do I receive?

You receive a PDF download immediately after purchase. Pro tier also includes a Notion template you can duplicate to your workspace. The PDF is formatted for easy browsing with a table of contents, category chapters, and copy-ready prompt blocks.

Can I share it with my team?

The kit is yours for personal use.

Do you update the prompts?

Yes. Depending on the size of the update, past buyers will receive generous discounts on the new version.

What if it doesn't work for me?

The prompts are shown in detail on this page before you buy — what you see is what you get. All sales are final.