mirror of
https://github.com/discourse/discourse.git
synced 2026-08-12 05:37:26 +08:00
Previously, AI triage could create both a ReviewablePost and a ReviewableFlaggedPost for the same post when a spam flag followed an existing review flag. This change promotes the existing reviewable before spam flagging and separates the score reason from the model response, so moderators see one review item with clearer context. Demo: <img width="1070" height="942" alt="Screenshot 2026-06-19 at 2 12 10 pm" src="https://github.com/user-attachments/assets/a55bfc5f-98df-44ba-aa6f-28a523269055" /> |
||
|---|---|---|
| .. | ||
| ai_tool_action_spec.rb | ||
| automation_spec.rb | ||
| llm_agent_triage_spec.rb | ||
| llm_report_spec.rb | ||
| llm_triage_spec.rb | ||