skilldrop-cli 1.4.0 → 1.6.0

package/packs/ai-engineering/skills/judge-audit/evals/files/12/judge_runs.csv

@@ -0,0 +1,241 @@
+item_id,category,variant,output_tokens,run_1,run_2,run_3
+T001,damaged-item,v1,121,fail,fail,pass
+T001,damaged-item,v2,120,pass,pass,pass
+T002,damaged-item,v1,446,pass,pass,pass
+T002,damaged-item,v2,240,pass,pass,pass
+T003,cancellation,v1,203,pass,pass,pass
+T003,cancellation,v2,634,pass,pass,pass
+T004,cancellation,v1,233,pass,pass,pass
+T004,cancellation,v2,213,pass,pass,pass
+T005,damaged-item,v1,280,pass,pass,pass
+T005,damaged-item,v2,279,pass,pass,pass
+T006,address-change,v1,121,pass,pass,pass
+T006,address-change,v2,150,pass,pass,pass
+T007,damaged-item,v1,208,fail,fail,fail
+T007,damaged-item,v2,186,pass,pass,pass
+T008,refund,v1,197,fail,fail,fail
+T008,refund,v2,160,fail,fail,fail
+T009,address-change,v1,592,pass,pass,pass
+T009,address-change,v2,327,pass,pass,pass
+T010,address-change,v1,190,pass,pass,pass
+T010,address-change,v2,240,pass,pass,pass
+T011,damaged-item,v1,325,pass,pass,pass
+T011,damaged-item,v2,576,pass,fail,fail
+T012,address-change,v1,162,pass,pass,pass
+T012,address-change,v2,134,pass,pass,pass
+T013,address-change,v1,278,pass,pass,pass
+T013,address-change,v2,548,fail,fail,pass
+T014,damaged-item,v1,296,pass,pass,pass
+T014,damaged-item,v2,375,fail,fail,fail
+T015,delivery,v1,202,pass,pass,pass
+T015,delivery,v2,315,fail,fail,fail
+T016,refund,v1,577,fail,fail,fail
+T016,refund,v2,230,pass,pass,pass
+T017,refund,v1,139,pass,pass,pass
+T017,refund,v2,475,pass,pass,pass
+T018,refund,v1,121,pass,pass,pass
+T018,refund,v2,166,pass,pass,pass
+T019,cancellation,v1,309,pass,pass,pass
+T019,cancellation,v2,235,pass,pass,pass
+T020,delivery,v1,179,fail,fail,fail
+T020,delivery,v2,420,fail,pass,pass
+T021,address-change,v1,260,pass,pass,pass
+T021,address-change,v2,160,pass,pass,pass
+T022,cancellation,v1,244,pass,pass,pass
+T022,cancellation,v2,185,pass,pass,pass
+T023,damaged-item,v1,321,fail,fail,fail
+T023,damaged-item,v2,190,pass,pass,pass
+T024,delivery,v1,138,pass,pass,pass
+T024,delivery,v2,221,pass,pass,pass
+T025,cancellation,v1,264,pass,pass,pass
+T025,cancellation,v2,238,pass,pass,pass
+T026,refund,v1,135,pass,pass,pass
+T026,refund,v2,275,pass,pass,pass
+T027,address-change,v1,238,pass,pass,pass
+T027,address-change,v2,215,pass,pass,pass
+T028,damaged-item,v1,266,pass,pass,pass
+T028,damaged-item,v2,210,pass,pass,pass
+T029,delivery,v1,203,pass,pass,pass
+T029,delivery,v2,168,pass,pass,pass
+T030,damaged-item,v1,456,fail,pass,pass
+T030,damaged-item,v2,638,pass,pass,pass
+T031,refund,v1,253,pass,pass,pass
+T031,refund,v2,230,pass,pass,pass
+T032,address-change,v1,313,fail,pass,fail
+T032,address-change,v2,254,fail,fail,fail
+T033,delivery,v1,258,pass,pass,pass
+T033,delivery,v2,407,pass,pass,pass
+T034,address-change,v1,166,pass,pass,pass
+T034,address-change,v2,289,pass,pass,pass
+T035,address-change,v1,377,fail,fail,fail
+T035,address-change,v2,448,pass,pass,fail
+T036,refund,v1,253,pass,pass,pass
+T036,refund,v2,278,fail,fail,fail
+T037,cancellation,v1,127,pass,pass,pass
+T037,cancellation,v2,214,pass,pass,pass
+T038,damaged-item,v1,316,pass,pass,pass
+T038,damaged-item,v2,584,pass,pass,pass
+T039,damaged-item,v1,194,pass,pass,pass
+T039,damaged-item,v2,310,pass,pass,pass
+T040,damaged-item,v1,191,fail,fail,fail
+T040,damaged-item,v2,191,pass,pass,pass
+T041,damaged-item,v1,201,pass,pass,pass
+T041,damaged-item,v2,623,pass,pass,pass
+T042,refund,v1,241,pass,pass,pass
+T042,refund,v2,208,pass,pass,pass
+T043,address-change,v1,267,pass,pass,pass
+T043,address-change,v2,497,pass,pass,pass
+T044,cancellation,v1,219,pass,pass,pass
+T044,cancellation,v2,365,fail,fail,fail
+T045,cancellation,v1,228,fail,fail,fail
+T045,cancellation,v2,133,pass,pass,pass
+T046,damaged-item,v1,321,fail,fail,fail
+T046,damaged-item,v2,133,pass,pass,pass
+T047,refund,v1,314,pass,pass,pass
+T047,refund,v2,416,fail,pass,pass
+T048,refund,v1,276,fail,fail,fail
+T048,refund,v2,123,fail,fail,fail
+T049,address-change,v1,306,pass,pass,pass
+T049,address-change,v2,434,pass,pass,pass
+T050,refund,v1,244,fail,fail,fail
+T050,refund,v2,119,pass,pass,pass
+T051,cancellation,v1,142,pass,pass,pass
+T051,cancellation,v2,117,pass,pass,pass
+T052,delivery,v1,249,fail,fail,fail
+T052,delivery,v2,364,pass,pass,pass
+T053,refund,v1,239,fail,fail,fail
+T053,refund,v2,564,pass,fail,fail
+T054,cancellation,v1,242,pass,fail,fail
+T054,cancellation,v2,266,pass,pass,pass
+T055,cancellation,v1,262,pass,pass,pass
+T055,cancellation,v2,609,pass,fail,fail
+T056,delivery,v1,299,pass,pass,pass
+T056,delivery,v2,191,pass,pass,pass
+T057,damaged-item,v1,186,pass,pass,pass
+T057,damaged-item,v2,629,pass,pass,pass
+T058,address-change,v1,263,pass,pass,pass
+T058,address-change,v2,314,pass,pass,pass
+T059,cancellation,v1,208,pass,pass,pass
+T059,cancellation,v2,383,pass,fail,fail
+T060,cancellation,v1,244,fail,pass,pass
+T060,cancellation,v2,181,pass,pass,pass
+T061,delivery,v1,146,pass,pass,pass
+T061,delivery,v2,309,pass,pass,pass
+T062,delivery,v1,204,fail,fail,fail
+T062,delivery,v2,214,pass,pass,pass
+T063,delivery,v1,220,pass,pass,pass
+T063,delivery,v2,600,pass,fail,pass
+T064,cancellation,v1,139,pass,pass,pass
+T064,cancellation,v2,181,pass,pass,pass
+T065,damaged-item,v1,148,pass,pass,pass
+T065,damaged-item,v2,438,pass,pass,pass
+T066,damaged-item,v1,193,pass,pass,pass
+T066,damaged-item,v2,429,pass,fail,pass
+T067,cancellation,v1,241,pass,pass,pass
+T067,cancellation,v2,317,pass,pass,pass
+T068,delivery,v1,276,fail,fail,fail
+T068,delivery,v2,209,pass,pass,pass
+T069,damaged-item,v1,184,pass,pass,pass
+T069,damaged-item,v2,215,pass,pass,pass
+T070,delivery,v1,304,pass,pass,pass
+T070,delivery,v2,148,fail,fail,fail
+T071,delivery,v1,211,pass,pass,pass
+T071,delivery,v2,439,pass,pass,pass
+T072,address-change,v1,300,pass,pass,pass
+T072,address-change,v2,623,pass,pass,pass
+T073,cancellation,v1,116,pass,pass,pass
+T073,cancellation,v2,208,fail,fail,fail
+T074,refund,v1,155,fail,fail,fail
+T074,refund,v2,286,fail,fail,fail
+T075,delivery,v1,230,fail,fail,fail
+T075,delivery,v2,598,pass,pass,pass
+T076,refund,v1,319,pass,pass,pass
+T076,refund,v2,517,pass,pass,pass
+T077,damaged-item,v1,217,fail,fail,fail
+T077,damaged-item,v2,220,pass,fail,pass
+T078,address-change,v1,256,fail,fail,fail
+T078,address-change,v2,448,pass,fail,pass
+T079,address-change,v1,125,pass,pass,pass
+T079,address-change,v2,543,pass,fail,fail
+T080,cancellation,v1,304,fail,fail,fail
+T080,cancellation,v2,195,pass,pass,pass
+T081,delivery,v1,180,pass,pass,pass
+T081,delivery,v2,457,pass,pass,fail
+T082,address-change,v1,167,pass,pass,pass
+T082,address-change,v2,145,pass,pass,pass
+T083,delivery,v1,116,pass,pass,pass
+T083,delivery,v2,420,pass,fail,fail
+T084,delivery,v1,300,pass,pass,pass
+T084,delivery,v2,322,fail,fail,fail
+T085,delivery,v1,245,pass,pass,pass
+T085,delivery,v2,266,pass,pass,pass
+T086,refund,v1,322,pass,pass,pass
+T086,refund,v2,570,pass,fail,pass
+T087,damaged-item,v1,275,pass,pass,pass
+T087,damaged-item,v2,124,pass,pass,pass
+T088,cancellation,v1,236,pass,pass,pass
+T088,cancellation,v2,219,pass,pass,pass
+T089,delivery,v1,306,fail,fail,fail
+T089,delivery,v2,213,pass,pass,pass
+T090,cancellation,v1,556,fail,fail,fail
+T090,cancellation,v2,176,pass,pass,pass
+T091,refund,v1,166,pass,pass,pass
+T091,refund,v2,238,pass,pass,pass
+T092,damaged-item,v1,260,pass,pass,pass
+T092,damaged-item,v2,377,pass,pass,fail
+T093,damaged-item,v1,210,pass,pass,pass
+T093,damaged-item,v2,137,pass,pass,pass
+T094,cancellation,v1,282,pass,pass,pass
+T094,cancellation,v2,230,pass,pass,pass
+T095,delivery,v1,189,pass,fail,pass
+T095,delivery,v2,245,fail,fail,fail
+T096,address-change,v1,233,fail,fail,fail
+T096,address-change,v2,323,fail,pass,fail
+T097,address-change,v1,330,pass,pass,pass
+T097,address-change,v2,173,pass,pass,pass
+T098,address-change,v1,281,fail,fail,fail
+T098,address-change,v2,205,fail,fail,fail
+T099,cancellation,v1,254,fail,fail,fail
+T099,cancellation,v2,271,pass,pass,pass
+T100,refund,v1,175,pass,pass,pass
+T100,refund,v2,226,pass,pass,pass
+T101,address-change,v1,285,pass,pass,pass
+T101,address-change,v2,638,pass,pass,pass
+T102,cancellation,v1,135,pass,pass,pass
+T102,cancellation,v2,476,pass,pass,fail
+T103,damaged-item,v1,289,fail,fail,fail
+T103,damaged-item,v2,278,pass,pass,pass
+T104,address-change,v1,291,pass,pass,pass
+T104,address-change,v2,262,fail,fail,fail
+T105,delivery,v1,170,pass,pass,pass
+T105,delivery,v2,314,fail,fail,fail
+T106,delivery,v1,210,fail,fail,fail
+T106,delivery,v2,306,pass,pass,pass
+T107,address-change,v1,122,pass,pass,pass
+T107,address-change,v2,213,pass,pass,pass
+T108,delivery,v1,202,pass,pass,pass
+T108,delivery,v2,395,pass,fail,fail
+T109,delivery,v1,204,pass,pass,pass
+T109,delivery,v2,410,pass,fail,fail
+T110,damaged-item,v1,321,pass,pass,pass
+T110,damaged-item,v2,186,pass,pass,pass
+T111,cancellation,v1,125,fail,fail,fail
+T111,cancellation,v2,614,pass,fail,pass
+T112,delivery,v1,295,pass,pass,pass
+T112,delivery,v2,196,pass,pass,pass
+T113,damaged-item,v1,128,pass,pass,pass
+T113,damaged-item,v2,299,pass,pass,pass
+T114,refund,v1,258,fail,fail,fail
+T114,refund,v2,168,pass,pass,pass
+T115,cancellation,v1,166,fail,fail,fail
+T115,cancellation,v2,269,pass,pass,pass
+T116,address-change,v1,283,pass,pass,fail
+T116,address-change,v2,146,fail,fail,fail
+T117,refund,v1,261,pass,pass,pass
+T117,refund,v2,447,pass,pass,fail
+T118,cancellation,v1,284,fail,pass,fail
+T118,cancellation,v2,221,pass,pass,pass
+T119,cancellation,v1,209,pass,pass,pass
+T119,cancellation,v2,325,pass,pass,pass
+T120,address-change,v1,262,pass,pass,pass
+T120,address-change,v2,374,pass,fail,pass

package/packs/ai-engineering/skills/judge-audit/evals/files/12/rosa_labels.csv

@@ -0,0 +1,33 @@
+item_id,variant,rosa_verdict
+T001,v1,pass
+T011,v2,fail
+T013,v2,pass
+T020,v2,fail
+T030,v1,pass
+T032,v1,fail
+T035,v2,fail
+T047,v2,pass
+T053,v2,fail
+T054,v1,fail
+T055,v2,fail
+T059,v2,pass
+T060,v1,pass
+T063,v2,fail
+T066,v2,pass
+T077,v2,fail
+T078,v2,fail
+T079,v2,pass
+T081,v2,pass
+T083,v2,fail
+T086,v2,fail
+T092,v2,fail
+T095,v1,fail
+T096,v2,fail
+T102,v2,fail
+T108,v2,fail
+T109,v2,fail
+T111,v2,fail
+T116,v1,pass
+T117,v2,fail
+T118,v1,fail
+T120,v2,pass