Can you explain what type of features you expect AI Mentor for Debugger should have?

Sure Daniël 

I imagine an AI Mentor in Debugger could assist in several ways:

• Smart error interpretation – Instead of just showing a runtime error or null reference, the AI could explain the likely cause in plain language (e.g., “This variable is null because the aggregate didn’t return results; consider handling an empty list”).

• Proactive fix suggestions – Propose possible corrections with reasoning (e.g., missing initialization, wrong data type mapping, forgotten ‘await’ in async server action).

• Pattern-based learning – Identify common pitfalls in OutSystems apps (like inefficient aggregates, redundant logic, or circular references) and suggest best practices.

• Root cause tracing – Not just the failing step, but highlighting the full dependency chain that led to the issue.

• Knowledge integration – Pull in related documentation, forum posts, or OutSystems patterns directly relevant to the error.

• Optimization insights – While debugging performance issues, the AI could recommend reducing loops, caching queries, or reusing data actions.

Bullet 1, 2 and 5 are more related to error logging not the debugger.

Bullet 3 are patterns already reported by AI Mentor Studio.

Bullet 4, isn't that the call stack on an error?