Responding empathetically

I see the user used a Chinese sentence that translates to, "This isn't a mistake; it's a disaster." They didn't ask a question, but it seems like they might be feeling frustrated. To assist them, I need to respond carefully, asking for context and showing empathy. I should suggest gathering facts and checking logs, perhaps offering commands related to their situation while keeping it general. I'll provide four options for next steps to keep it concise and supportive.Outlining immediate actions

I need to provide immediate actions to help manage the situation. This includes stopping the bleeding, disabling feature flags, rolling back if necessary, cutting traffic, and locking deployments. Then, I'll suggest gathering important info like snapshots, logs, and timestamps. A recovery plan could involve restoring from backups, rolling back releases, applying hotfixes, and maintaining communication. I can ask clarifying questions like, “What’s the context (code, infrastructure, data)?” and share a simple summary template. I’ll also suggest relevant git commands for a hotfix without running any shell commands just yet.Crafting a response

I want to start by showing empathy, saying something like, "Don't blame yourself. Let's stop this, assess, and fix it together." Then, I should ask for details by providing four specific questions. I need to include a checklist for immediate containment, outlining quick rollback steps if the outage is due to a code deploy. If it’s a data mistake, I’ll suggest stopping writers and taking snapshots. I’ll keep lists concise with 4-6 bullets and a "Pick one" list for the user to choose from. Since the user is Chinese, I’ll respond in Chinese while maintaining a friendly tone.Providing step-by-step guidance

I’ll structure my response by giving clear actions depending on the type of incident. If it’s an online issue, I’ll suggest freezing the release, rolling back to a stable version, disabling new features, and implementing rate limiting while marking time points for logging. For a data incident, I’ll recommend halting writes, taking snapshots of original data, finding the latest reliable backup, and validating recovery in an isolated environment.

If it’s a local code issue, I’ll ask for git status and recent commit logs. Then, I’ll follow up with, “Tell me which category it falls under (1-4) or share the errors/screenshots/log lines, and I’ll provide specific commands or recovery steps.” I can include short rollback commands as well, keeping each list to 4-6 concise bullets.