AI Debugging Assistant
Give Mother the error, context, expected behavior, and constraints. Ask for root cause, smallest safe fix, and verification.
Start this task in Mother →No account required. The example opens in your composer; edit it before sending. Use uncensored AI for useful work with fewer content restrictions, not a guarantee of an answer.
What to provide
Describe the observable failure before proposing a fix. Include the exact error, expected behavior, actual behavior, a minimal reproduction and the first point where results differ. Name your runtime and any recent change. Reduce logs to the relevant lines and remove credentials or personal data. Ask for competing explanations, a check that distinguishes them and the smallest reversible change. If a symptom occurs only sometimes, include what varies between successful and failed attempts rather than presenting one guess as the cause.
Example starting request
Help me with debugging. Give Mother the error, context, expected behavior, and constraints. Ask for root cause, smallest safe fix, and verification. Explain your reasoning and identify anything I should verify.
Worked example
Illustrative example, not a generated result or a user transcript.
Debugging brief: “A JavaScript filter should keep records whose active field is the boolean true. It also keeps a record where active is the string "false". The predicate is item => item.active. Diagnose the mismatch and propose a regression test.” Illustrative diagnosis: a nonempty string is truthy, so the predicate does not check the stated boolean contract. A strict comparison with true would distinguish the two supplied cases.
The reproduction narrows the problem to value types instead of blaming the browser or suggesting a dependency upgrade. Keep a failing test for the string value, plus tests for boolean true, boolean false and a missing field. If the upstream contract allows string values, fix or normalize that boundary deliberately rather than quietly changing the filter. This is a hypothetical diagnosis grounded in the supplied example, not evidence about your real application. Confirm the actual data before applying the same fix.
Check the answer
Prove the reproduction fails before changing code and passes afterward. Run neighboring paths that share the same parser or predicate; otherwise the same flaw can survive elsewhere. Check that the fix preserves valid inputs and does not suppress errors or weaken validation just to make the symptom disappear. Retain the original failure log and compare the exact changed behavior. Do not accept “probably fixed” on the strength of explanation alone. When evidence is missing, ask for the next diagnostic observation instead of stacking speculative patches.