{"id":"cx-sla-breach-causes-false-breach-count","audience":"operations","category":"support-work","technique":"verification","track":"support","level":"intermediate","title":"Count real SLA breach causes and separate the false breaches","free":false,"task":"Start with the line \"BREACH REVIEW - week 44\". Then output a table with the header \"Cause | Tickets\" listing only real causes, most tickets first, then two last lines \"False breaches: N\" and \"Top cause: X\". A breach caused by a pending-status error is a false breach and is not a real cause. Plain text, under 50 words.","materialLabel":"Your material (attached to your prompt automatically)","material":"BREACHED TICKETS THIS WEEK (cause as tagged by the team lead)\nB1: Routing (sent to the wrong queue)\nB2: Pending-status error (clock kept running while waiting on the customer)\nB3: Staffing (no agent available)\nB4: Staffing\nB5: Routing\nB6: Staffing\nB7: Pending-status error\nB8: Complex case (needed engineering)\n\nNOTES\nPending-status errors are a dashboard problem. Per policy, the clock pauses while waiting on the customer, so those tickets did not truly breach. The team lead tagged the causes on Friday from the ticket history. Engineering is fixing the dashboard setting next sprint, and until then the weekly report has to correct the numbers by hand. Week 44 ran from Oct 26 to Nov 1 and covered 410 closed tickets.","rules":[{"type":"plainText"},{"type":"maxWords","max":50}],"mistakes":[{"title":"Counting false breaches as a cause","why":"The AI lists Pending-status error with 2.","fix":"Say: \"Pending-status errors are false breaches, not causes.\"","checks":["r1","r2"]},{"title":"Wrong order","why":"The AI keeps the ticket order.","fix":"Say: \"Sort by tickets, most first.\"","checks":["r1"]},{"title":"Missing literal lines","why":"The AI skips the title or last lines.","fix":"Quote the title line and the two last lines.","checks":["r0","r2"]},{"title":"Getting markdown back","why":"AI adds bold and headings by default, which show up as stray symbols where this is pasted.","fix":"Say \"plain text, no bold, no headings, no asterisks\".","checks":["plainText"]},{"title":"Trusting the AI to count words","why":"AI is bad at counting and usually runs long, so a vague \"keep it short\" lands well over the limit.","fix":"Cap each part instead of the total, for example \"each line under 15 words\".","checks":["maxWords"]}],"promptHints":[{"label":"Quotes the title","patterns":["breach review - week 44"]},{"label":"Gives the table header","patterns":["cause \\| tickets"]},{"label":"Explains false breaches","patterns":["false breach","pending"]},{"label":"Asks for plain text","patterns":["plain","no bold","no markdown","no formatting","no headings"]},{"label":"Sets a length limit","patterns":["\\d+\\s*words","word (limit|count)","under \\d+"]}],"hint":"Quote the title, give the header, and say how pending-status errors are handled.","rubricLabels":["The first line is exactly \"BREACH REVIEW - week 44\" and the next line is the header \"Cause | Tickets\"","The rows are Staffing 3, Routing 2, Complex case 1 in that order, with no row for the pending-status error","The last two lines are exactly \"False breaches: 2\" and \"Top cause: Staffing\""]}