design blocks 105, completed 105, missing 0

=== reader claude-opus-5-5: 63 blocks, 21 problems, infra-touched blocks 0, P1 failed (not correct or <1x) 9 | metric log1p
arm means (mean over problems of per-problem rep means): bare 1.084, show_all 0.911, bare2 1.337, attempt_then_consult 1.242, consult_on_failure 1.317, audit_then_consult 1.287, prior_then_consult 1.325
  PRIMARY attempt_then_consult - bare2             -0.095 [-0.190, -0.008]  problems up/down 4/17
  audit_then_consult - bare2                       -0.051 [-0.120, +0.019]  problems up/down 8/13
  consult_on_failure - bare2                       -0.020 [-0.044, -0.003]  problems up/down 0/4
  attempt_then_consult - prior_then_consult        -0.083 [-0.161, -0.021]  problems up/down 6/15
  audit_then_consult - attempt_then_consult        +0.044 [-0.002, +0.092]  problems up/down 14/6
  prior_then_consult - bare2                       -0.012 [-0.100, +0.058]  problems up/down 8/13
  show_all - bare                                  -0.173 [-0.343, -0.020]  problems up/down 6/15
  bare2 - bare                                     +0.253 [+0.144, +0.377]  problems up/down 18/1
  attempt_then_consult - bare                      +0.158 [+0.087, +0.244]  problems up/down 17/3
  audit_then_consult - bare                        +0.202 [+0.121, +0.289]  problems up/down 18/1
  consult_on_failure - bare                        +0.233 [+0.133, +0.346]  problems up/down 18/1
  prior_then_consult - bare                        +0.241 [+0.145, +0.353]  problems up/down 19/2
  attempt_then_consult - show_all                  +0.331 [+0.202, +0.492]  problems up/down 20/1
  consult_on_failure - prior_then_consult          -0.008 [-0.077, +0.078]  problems up/down 10/11
  audit_then_consult - prior_then_consult          -0.038 [-0.104, +0.013]  problems up/down 10/11
  diag P2_consult - P2_refine (raw second pass)    -0.061 [-0.183, +0.048]  n_problems 21
  diag P2_audit - P2_refine (raw second pass)      -0.022 [-0.107, +0.063]  n_problems 21
  diag P2_prior - P2_refine (raw second pass)      +0.011 [-0.089, +0.095]  n_problems 21
  diag S - P1 (raw pass evals)                     -0.171 [-0.341, -0.019]  n_problems 21
  selection (which pass each two-pass arm kept): bare2: {'P1': 21, 'P2_refine': 42}; attempt_then_consult: {'P1': 19, 'P2_consult': 44}; audit_then_consult: {'P2_audit': 47, 'P1': 16}; prior_then_consult: {'P1': 14, 'P2_prior': 49}

=== reader claude-opus-4-8: 42 blocks, 21 problems, infra-touched blocks 0, P1 failed (not correct or <1x) 31 | metric log1p
arm means (mean over problems of per-problem rep means): bare 0.389, show_all 0.701, bare2 0.706, attempt_then_consult 0.782, consult_on_failure 0.785, audit_then_consult 0.779, prior_then_consult 0.737
  PRIMARY attempt_then_consult - bare2             +0.075 [-0.021, +0.228]  problems up/down 12/9
  audit_then_consult - bare2                       +0.073 [-0.029, +0.230]  problems up/down 13/8
  consult_on_failure - bare2                       +0.078 [-0.017, +0.235]  problems up/down 13/7
  attempt_then_consult - prior_then_consult        +0.045 [-0.009, +0.114]  problems up/down 12/9
  audit_then_consult - attempt_then_consult        -0.003 [-0.031, +0.022]  problems up/down 12/8
  prior_then_consult - bare2                       +0.031 [-0.024, +0.122]  problems up/down 9/12
  show_all - bare                                  +0.312 [+0.159, +0.456]  problems up/down 17/4
  bare2 - bare                                     +0.317 [+0.212, +0.430]  problems up/down 18/2
  attempt_then_consult - bare                      +0.392 [+0.257, +0.547]  problems up/down 20/1
  audit_then_consult - bare                        +0.390 [+0.248, +0.547]  problems up/down 20/1
  consult_on_failure - bare                        +0.395 [+0.256, +0.549]  problems up/down 20/1
  prior_then_consult - bare                        +0.348 [+0.233, +0.468]  problems up/down 18/1
  attempt_then_consult - show_all                  +0.081 [+0.010, +0.166]  problems up/down 12/9
  consult_on_failure - prior_then_consult          +0.048 [-0.005, +0.116]  problems up/down 13/8
  audit_then_consult - prior_then_consult          +0.042 [-0.016, +0.113]  problems up/down 14/7
  diag P2_consult - P2_refine (raw second pass)    +0.114 [-0.004, +0.277]  n_problems 21
  diag P2_audit - P2_refine (raw second pass)      +0.075 [-0.073, +0.249]  n_problems 21
  diag P2_prior - P2_refine (raw second pass)      +0.037 [-0.062, +0.159]  n_problems 21
  diag S - P1 (raw pass evals)                     +0.284 [+0.134, +0.430]  n_problems 21
  selection (which pass each two-pass arm kept): bare2: {'P1': 12, 'P2_refine': 30}; attempt_then_consult: {'P1': 8, 'P2_consult': 34}; audit_then_consult: {'P1': 10, 'P2_audit': 32}; prior_then_consult: {'P1': 16, 'P2_prior': 26}

(secondary diagnostic: metric log1p, wave pooled; preregistered decision rules apply only to --wave 1 --metric score and are not evaluated here)