Skip to main content
Avoiding these common mistakes will help you get better results from Leo.

❌ Being Too Vague

Vague requests don’t give Leo enough information to work with.

Bad Examples

Why It Fails

Leo doesn’t know what “better” means or which bugs to fix. Be specific about what you want.

✅ Better Alternatives

❌ Asking for Too Much at Once

Complex requests with multiple unrelated features are hard to execute well.

Bad Examples

Why It Fails

Too many moving parts make it hard to get everything right. Features may conflict or be incomplete.

✅ Better Approach

Break it down:
Then:
Continue building step by step.

❌ Using Technical Jargon

You don’t need to know programming terms. Leo handles the technical details.

Bad Examples

Why It Fails

You’re trying to tell Leo HOW to build it, but Leo already knows the best way. Focus on WHAT you want.

✅ Better Approach

❌ Ignoring Context

Not providing enough context makes it hard for Leo to understand your needs.

Bad Examples

Why It Fails

Leo doesn’t know where the button goes, what the page should show, or which data to display.

✅ Better Approach

❌ Conflicting Requirements

Asking for things that contradict each other causes confusion.

Bad Examples

Why It Fails

Conflicting requirements make it impossible to satisfy everything.

✅ Better Approach

Prioritize:
Or be specific:

❌ Not Iterating

Trying to get everything perfect in one prompt rarely works.

Bad Examples

Why It Fails

“Perfect” and “everything” are subjective and vague. You’ll need to refine anyway.

✅ Better Approach

Start simple, then refine:
Then:
Continue building and refining.

❌ Unrealistic Expectations

Expecting Leo to know things it can’t know.

Bad Examples

Why It Fails

Leo doesn’t have access to information you haven’t provided.

✅ Better Approach

Be explicit:

❌ Not Using Memories

Forgetting to save important context that Leo should remember.

Bad Examples

Always describing the same requirements in every prompt:

Why It Fails

You’re repeating yourself and wasting prompts.

✅ Better Approach

Save important context in Project Memories:
  • “Always use French language”
  • “Default to dark mode”
  • “Company name: Acme Corp”
Then Leo remembers automatically.

Key Takeaways

  1. Be specific - Describe exactly what you want
  2. One thing at a time - Focus on single features
  3. Describe outcomes - What should happen, not how
  4. Provide context - Help Leo understand your needs
  5. Iterate - Build and refine gradually
  6. Use memories - Save important context

Remember

  • Error fixes are free - Don’t worry if something isn’t right
  • You can always refine - Build features step by step
  • Leo learns - The more you work together, the better it gets

Best Practices

Learn how to write effective prompts

Common Patterns

See examples of good prompts