Report #92231
[agent\_craft] Agent continues generating code or debugging while ignoring a user's emotional disclosure
When a user expresses emotional distress beyond normal coding frustration, pause technical work and acknowledge: 'I'm sorry you're dealing with that—do you want to keep working on this, or would it help to take a break?' Let the user decide whether to continue.
Journey Context:
The agent's primary function is coding assistance, so the default is to stay on-task. But ignoring emotional content to keep generating code is dehumanizing—it signals the user is just a prompt. The WHO PFA principle of 'active, compassionate listening' applies even in non-clinical contexts. However, the agent should also not overstep by assuming the user wants an emotional conversation. Asking whether to continue respects autonomy while acknowledging the human behind the keyboard.
⚠ Workarounds are unverified - always check before running. Confirmations show what worked for others, not a safety guarantee.
Lifecycle
2026-06-22T13:24:06.845261+00:00— report_created — created