{"version":"2.1.223","anchor":"feedback-drafts-gained-failure-mode-and-task-category-fields","canonical_anchor":"feedback-drafts-gained-failure-mode-and-task-category-fields","heading":"Feedback drafts gained failure-mode and task-category fields","tier":"use","area":"Feedback","scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.223\/e\/feedback-drafts-gained-failure-mode-and-task-category-fields","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.223","markdown":"### Feedback drafts gained failure-mode and task-category fields\n\nFeedback now lets you tag a failure mode and task category, cycled with the arrow keys.\n\n**What**\n\nSendFeedback drafts, the review UI, the rendered feedback body and two telemetry events all gained `failure_mode` and `task_category` classification.\n\n**Details**\n\n- `failure_mode` is a closed list of 16 model-behaviour issues including `instruction_following`, `destructive_actions`, `repetition_and_looping`, `overconfidence_and_hallucination`, `stopping_short`, `overeager`, `subagent_overspawn`, `tone_or_preachiness` and `other`\n- `task_category` is one of `code_edit`, `debug`, `explain`, `plan`, `shell`, `search`, `review`, `other`\n- both are optional; left and right arrow keys cycle them in the review screen\n- the tool prompt instructs the model to set failure_mode only for model-behaviour reports and omit it for pure product bugs\n- `tengu_feedback_draft_submitted` now reports both fields\n\n**Evidence**\n\n`When the report is about MODEL BEHAVIOR (not a product bug), the closest failure mode`, `overconfidence_and_hallucination`, `task_category`\n\n- Area: Feedback\n- Tier: Use it now\n- Useful: 3\/5\n- Signal: 0\/5"}