A coding agent writes a plausible fix and breaks adjacent behavior. How would you catch that?

Instruction: Describe how you would protect against locally correct but globally harmful patches. Python contract: an empty sequence sums to 0; a nonempty numeric sequence returns its total, including negatives. Proposed patch is return max(sum(xs), 0) to fix a UI that displayed -0.0. Give a counterexample, diagnosis and narrower next step.

Context: Tests how the candidate diagnoses the problem, chooses the safest next step, and reasons through recovery. Describe how you would protect against locally correct but globally harmful patches.

Updated

Official answer available

Read the opening below, then unlock the full answer and practical guidance.

I'd reproduce the reported failure, then trace callers and shared behavior before accepting the fix...

Upgrade to view official answer

Your preparation path

Work through these questions in order. Read the answer aloud, then explain it in your own words.

1. Start with the foundations

Build the vocabulary and explain the core decisions.

  1. When is a coding agent the right tool instead of autocomplete or manual engineering? Free sample
  2. What information should a coding agent persist between steps? Free sample
  3. Why do tests matter differently for coding agents than for chatbots? Free sample

2. Apply it to a real workflow

Practice diagnosis, validation, and everyday tradeoffs.

  1. What makes a patch proposal safe enough to apply automatically? Free sample
  2. The agent can find the bug, but its patches are hard for humans to review. What would you change? Member answer
  3. How would you verify a coding agent's web feature beyond a passing build and a screenshot? Free sample

3. Prepare for senior discussions

Explain failure boundaries, recovery, and production choices.

  1. A coding agent can modify infrastructure, app code, and data access layers in one task. How do you constrain it? Member answer
  2. Two coding agents change an API and its client in parallel. How would you integrate their work safely? Member answer
  3. A coding-agent hook runs before the repository trust prompt. What is wrong with that design? Member answer

Related Questions