← All Cheatsheets

AI for Developers — Prompt Toolkit

Code Review · Documentation · Testing · Debugging · Refactoring · Architecture
mitraaiprojects.com

Code Review

You are a senior Python developer.
Review this code for:
1. Security vulnerabilities
2. Performance bottlenecks
3. Readability issues
4. Missing error handling
5. Potential bugs

Rate each 1-5 (5=no issues).
Provide specific line references.

```python
{paste_code_here}
```

Generate Tests

Write pytest unit tests for this function.
Cover:
- Happy path (normal inputs)
- Edge cases (empty, None, boundaries)
- Error cases (invalid inputs, exceptions)
- At least 8 test functions

```python
{function_code}
```

Use descriptive test names:
def test_should_return_X_when_Y():

Write Documentation

Write a Google-style docstring for this
function. Include:
- Brief one-line summary
- Args: each parameter with type + description
- Returns: type + description
- Raises: exceptions with conditions
- Example: working usage example

```python
{function_code}
```

Debug This Error

I'm getting this error in Python:

ERROR:
{paste_error_traceback}

CODE (relevant section):
```python
{paste_code}
```

CONTEXT:
- Input: {what_data_you_passed}
- Expected: {what_should_happen}
- Actual: {what_happened}

Please:
1. Explain what caused the error
2. Provide the fixed code
3. Explain how to prevent it

Refactor Code

Refactor this code to:
- Follow PEP 8 style guidelines
- Use appropriate design patterns
- Improve readability
- Add type hints
- Remove code duplication
- Improve performance where possible

Explain each significant change.
Keep the same public API.

```python
{paste_code}
```

Explain Code

Explain this code to a junior developer
who knows Python basics but has never
seen this pattern.

For each section:
1. What it does (plain English)
2. Why it's written this way
3. What would break if removed

```python
{paste_code}
```

System Design

Design a system for: {description}

Requirements:
- Scale: {users} users, {requests} req/s
- Latency: {target} p99
- Availability: {uptime}%

Please cover:
1. Architecture diagram (ASCII)
2. Component choices + reasons
3. Database schema (if applicable)
4. API design (key endpoints)
5. Scaling strategy
6. Failure modes + mitigations

SQL / Data Queries

Natural language → SQL
I have these tables:
{paste_schema_or_describe_tables}

Write a SQL query that:
{describe_what_you_want}

Optimise for performance.
Explain the query logic.

Debug slow query
This query is slow (~5s on 1M rows).
Suggest indexes and query optimisations.
```sql
{paste_slow_query}
```

ML Code Helpers

Implement from scratch:
Implement {algorithm} in Python
without using sklearn.
Include: forward pass, loss, gradients.

Optimise training loop:
This training loop is slow.
Add: mixed precision, gradient
checkpointing, pin_memory.
```python
{paste_training_code}
```

Debug model not converging:
My model loss is {behaviour}.
Architecture: {describe_model}
Hyperparameters: {list_params}
What could cause this? How to fix?

Git & DevOps Prompts

Write commit message:
Write a conventional commit message for:
Changed: {describe_changes}
Reason: {why_changed}
Format: type(scope): description

Write PR description:
Write a PR description for these changes:
{git_diff_or_description}
Include: what changed, why, how tested.

Write Dockerfile:
Write an optimised Dockerfile for a
Python {framework} app with:
- Non-root user for security
- Efficient layer caching
- Health check endpoint
- requirements.txt already present

Power Tips

TipHow
Get shorter answers"Reply in under 100 words"
Get code only"Return ONLY the code, no explanation"
Continue generation"Continue from where you left off"
Force specific formatShow an example of the format you want
Reduce hallucination"If you're not sure, say so explicitly"
Iterative refinement"Now make it more efficient" / "Now add error handling"
Different angle"Approach this differently, focusing on X"
Role assignment"Act as a {expert} reviewing this {artifact}"