Introduce PromptAi support: add model, migration, seeder and controller with full CRUD endpoints, and register routes under /api/prompt-ai. Integrate DB-driven prompts into LineConnection (replace hardcoded prompt strings for DPELP and ENV with fetched PromptAi records). Update frontend ModalConfig to add a Prompt AI management tab (fetch, create, edit prompts), plus related UI tweaks (tabs, prompt editor modal, axios/notifications). This makes AI prompts editable at runtime without code changes.
Increase physical test polling interval from 5s to 15s (backend setInterval and frontend AutoProgress ms) to reduce polling frequency. Add handling to stop the interval when physicalTest.done, send the physical test report, and end testing. Change license filtering to be case-insensitive and match any value containing "permanent" (backend and frontend) so licenses are detected regardless of casing/format.
Emit a dedicated 'update_status_ready' event when a LineConnection becomes ready, and stop sending isReady in the 'line_output' payload. On the server side, normalize incoming run_all_dpelp data and filter lineIds to only include lines that are already ready; also bail out early if no eligible lines or station isn't configured to send wiki. On the frontend, initialize line state with isReady, stop deriving readiness from line_output, and listen for the new 'update_status_ready' event to update line readiness in the UI.
Reset latestScenario when a line disconnects and surface AI-detected issues in the UI. Backend: clear config.latestScenario on disconnect. Frontend: initialize latestScenario in App state, import useEffect and add isShowIssue state in CardLine with an effect that checks line.latestScenario.detectAI.issue (array) to toggle the badge. UI: reorganize feature badges layout and render a red "Issue" pill when an AI-detected issue exists.
andrew.ng
joseph le
nguyentrungthat