What happens when a query fails or errors in iDBQuery?
When a query errors — a bad column reference, a type mismatch, an unreachable source — iDBQuery reads the database error, corrects the SQL, and retries automatically, often resolving it without you noticing. If it genuinely can't complete, it tells you plainly what went wrong instead of returning a made-up number.
Errors are normal when generating SQL against real schemas, and how a tool recovers matters more than whether an error ever occurs.
- Self-correction. If the database rejects a query — an unknown column, a wrong type, a syntax slip — iDBQuery reads the error message, adjusts the SQL, and tries again. This loop resolves many issues silently.
- Honest failure. When it truly can't answer — a source is unreachable, the data needed doesn't exist, or the question is out of scope — it says so clearly rather than inventing a plausible-looking figure.
- Helpful next step. Often it explains why and suggests a reframing: pointing at the right table, narrowing a date range, or asking you to clarify a term.
This reflects iDBQuery's core commitment to trustworthy answers over confident guesses. A wrong number quietly returned is worse than an honest "I couldn't determine that," so iDBQuery is built to prefer the latter. And because successful answers come with visible SQL and row-level citations, you can always distinguish a solid result from a case where iDBQuery has flagged uncertainty and asked for more information.
Updated 2026-06-22