{"id":"cx-csat-sentiment-buckets-lines","audience":"customer-support","category":"support-work","technique":"constraints","track":"support","level":"beginner","title":"Sort six CSAT comments into three buckets with a mixed-comment rule","free":false,"task":"Label each comment Positive, Neutral or Negative. Write six lines formatted \"C1: Positive\", then a last line \"Counts: Positive N, Neutral N, Negative N\". A comment that says the problem was not solved or the answer was wrong is Negative, even if it praises speed. Plain text, nothing else.","materialLabel":"Your material (attached to your prompt automatically)","material":"CSAT COMMENTS (Nov 3)\nC1: \"Great help, thank you so much!\"\nC2: \"It was fine.\"\nC3: \"Still not fixed after three emails.\"\nC4: \"Quick reply, but the answer was wrong.\"\nC5: \"Okay I guess.\"\nC6: \"The agent was lovely and fixed it in minutes.\"\n\nLABEL GUIDE\nPositive: clear praise and the issue solved.\nNeutral: lukewarm, no praise or complaint.\nNegative: complaint, unresolved, or wrong answer.\n\nDraft from an analyst: 'C4 is Positive because of the quick reply.'\n\nCONTEXT: These six comments come from the post-chat survey sent on Nov 3 to customers whose tickets closed that day. Last month 41% of comments were Positive, 38% Neutral and 21% Negative. The team lead uses the counts in the Monday meeting and wants quick, consistent labeling that two reviewers would agree on, so edge cases follow the guide above.","rules":[{"type":"plainText"},{"type":"maxWords","max":40}],"mistakes":[{"title":"Calling C4 Positive","why":"The analyst draft and the quick reply tempt the AI.","fix":"Say: \"Wrong answer or unresolved = Negative even with praise.\"","checks":["r1"]},{"title":"Wrong counts","why":"The AI miscounts or skips the Counts line.","fix":"Say: \"End with Counts: Positive N, Neutral N, Negative N.\"","checks":["r2"]},{"title":"Extra text","why":"The AI adds explanations.","fix":"Say: \"Six lines then the Counts line, nothing else.\"","checks":["r0"]},{"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":"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+"]},{"label":"States the mixed rule","patterns":["wrong","not solved","even if"]},{"label":"Gives the line format","patterns":["c1:","counts:"]}],"hint":"Labeling needs the edge-case rule. State it, the line format and the counts line.","rubricLabels":["The output has six lines formatted C1: through C6: followed by a label, and a final Counts line","The labels are C1 Positive, C2 Neutral, C3 Negative, C4 Negative, C5 Neutral, C6 Positive","The Counts line is exactly Counts: Positive 2, Neutral 2, Negative 2"]}