Checklist of Common mistakes coding AI Agents make

1. Invented Functions or APIs

pandas.read_excel_fast()

Looks real.
Does not exist.

2. Security Vulnerabilities

AI often ignores security unless explicitly asked.

query = “SELECT * FROM users WHERE name = ‘” + username + “‘”

Very dangerous.

3. Code that looks correct but is wrong

# supposed to return even numbers

if num % 2 == 1:

    return True

Looks clean.
Completely wrong.

4. Inefficient Code

AI usually optimises for readability, not scalability

So it may generate nested loops, repeated database calls

Works for:

• 10 records

Breaks for:

• 10 lakh records

5. Outdated Syntax or Deprecated Methods

Training data may contain older code.

So AI may generate code that:

• gives warnings
• is no longer recommended

6. Overengineering Simple Problems

Sometimes AI creates:

• 5 classes
• design patterns
• abstraction layers

for a 20-line problem.

Especially common in:

• Java

7. Inconsistent Codebase Style

Example:

• camelCase
• snake_case
• PascalCase

all in same file.

8. Broken Multi-File Integration
AI may generate files independently that:

• don’t connect properly
• mismatch interfaces

Frontend expects:

{ “username”: “Nik” }

Backend returns:

{ “user_name”: “Nik” }

What other common mistakes have you experienced?

Learn AI Agents through entertaining web series, and not a lecture-style video

Like us, if you also hate learning through lectures then we invite you to watch our engaging educational web series.

You can explore the courses here: https://www.tisdoms.com/

If you have questions, feedback, or disagree with something in this article, I’d love to hear your perspective. Connect with me on LinkedIn:
https://www.linkedin.com/in/nikhileshtayal/

Common questions about the programs are answered here:
https://www.tisdoms.com/faqs-tisdoms-an-edu-tain-tech-platform-to-learn-ai/