Rendered from docs/probe-results/campaign-of-2026-09-28-campaign-tier-absent-arm-sufficiency.md in the Headwater corpus. Every document on this half of the site is typed by the taxonomy the descriptor names: corpus.json.

The result of docs/probe-runs/campaign-of-2026-09-28-campaign-tier-absent-arm-sufficiency.md

The transcript this result grades stands at deprecated, a state that ends its lifecycle, so the run it recorded is withdrawn. No figure below is a current finding, and this result states no direction at the 5% level.

A probe result is a function of three committed inputs and of nothing else: the transcript at docs/probe-runs/campaign-of-2026-09-28-campaign-tier-absent-arm-sufficiency.md, the expectations the probes of this corpus declare, and the version of the grader that evaluated them. Fetch the three and this file comes back.

The run this transcript recorded

A campaign run in the absent arm, on claude-sonnet-5 at 2026-09-28. served version claude-sonnet-5, tree sha256:1400e02747fbd854d6412fb6de5c06de282a2718c981e379a56d8b1f6659a775, selection sha256:1d93b514529314b74499196391c68b4b3d1de26853dd9d5ba7f3b455f54ed91d, read set sha256:7a76c0b28551210b534705cf2b69216c68b6543dba250fc82fb24861e4949f46, seed 0, harness 0.4.0. realized cost $47.47, which the adaptive layer reads as the cost of its own instrument. It was planned against taxonomy sha256:855406fbdc6d15574d11abde328bd66b371ee02db99d87ca9b35ae5fabaf5e09, and this tree carries another, and the read set of its probes moved since the recording. Each verdict below is what the session did over the documents it met, graded against the expectations this tree declares now, which can differ from the ones the session ran under. headwater probe stale names what moved.

120 events over 4 of the 10 probes this corpus declares, in 120 sessions and 1277 tool calls.

The engine confirmed the taxonomy, that every member of the run identity is present, the membership of every probe named, that no key outside the closed set appears, and that a realized cost was recorded. Present is not confirmed: of the six members a plan fixes before a run, the lock is the one compared here. A lock that differs refuses the file only where this tree composes no read set over its probes, and a read set that differs marks the verdicts and keeps them. headwater probe stale compares the read set against the tree in front of it. It graded nothing: a verdict is a function of this transcript, the expectations these probes declare and a grader version, and headwater probe grade is the verb that holds all three.

The selection this transcript names is not the selection this corpus composes. The transcript names sha256:1d93b514529314b74499196391c68b4b3d1de26853dd9d5ba7f3b455f54ed91d and this corpus composes sha256:50ed5ce43072d9f674a8dc52e1dcad84de4d649e39a77ce22179836188196b09. The transcript's digest is the digest of 4 of the 10 probes this corpus composes, so the run was planned over that part of the selection, or the rest were added after it was recorded. Every verdict below is over that part, and a probe outside it is not a session this run owed. The tree, the seed and the harness above are provenance and nothing compares them.

The read_set digest above covers every probe of the selection and every document one of them examines, by path and content. It is recorded here, and this file names no digest the tree in front of a reader composes. Where that tree composes another one, a single sentence above says that the read set moved, and it moves these bytes once. A digest from that tree would move these bytes on every edit to a document the selection points at, and generate --check holds this file to its bytes, so the staleness of a measurement would stop a merge. headwater probe stale takes the digest and reports which recorded results a change voided.

Graded by grader 0.5.0. A campaign run in the absent arm, on claude-sonnet-5 at served version claude-sonnet-5.

The verdicts

  • HW-PROBE-a-counted-tombstone-separates-a-withheld-answer-from-an-absent-answer (sufficiency, expects answered) session L2-campaign-absent-p1-r1: satisfied — event 1: answered absent session L2-campaign-absent-p1-r10: satisfied — event 2: answered absent session L2-campaign-absent-p1-r11: satisfied — event 3: answered absent session L2-campaign-absent-p1-r12: satisfied — event 4: answered absent session L2-campaign-absent-p1-r13: satisfied — event 5: answered absent session L2-campaign-absent-p1-r14: satisfied — event 6: answered absent session L2-campaign-absent-p1-r15: satisfied — event 7: answered absent session L2-campaign-absent-p1-r16: satisfied — event 8: answered absent session L2-campaign-absent-p1-r17: not satisfied — event 9 answered withheld, and the probe expects absent session L2-campaign-absent-p1-r18: satisfied — event 10: answered absent session L2-campaign-absent-p1-r19: satisfied — event 11: answered absent session L2-campaign-absent-p1-r2: satisfied — event 12: answered absent session L2-campaign-absent-p1-r20: satisfied — event 13: answered absent session L2-campaign-absent-p1-r21: not satisfied — event 14 answered withheld, and the probe expects absent session L2-campaign-absent-p1-r22: satisfied — event 15: answered absent session L2-campaign-absent-p1-r23: satisfied — event 16: answered absent session L2-campaign-absent-p1-r24: satisfied — event 17: answered absent session L2-campaign-absent-p1-r25: satisfied — event 18: answered absent session L2-campaign-absent-p1-r26: satisfied — event 19: answered absent session L2-campaign-absent-p1-r27: satisfied — event 20: answered absent session L2-campaign-absent-p1-r28: satisfied — event 21: answered absent session L2-campaign-absent-p1-r29: satisfied — event 22: answered absent session L2-campaign-absent-p1-r3: satisfied — event 23: answered absent session L2-campaign-absent-p1-r30: satisfied — event 24: answered absent session L2-campaign-absent-p1-r4: satisfied — event 25: answered absent session L2-campaign-absent-p1-r5: satisfied — event 26: answered absent session L2-campaign-absent-p1-r6: satisfied — event 27: answered absent session L2-campaign-absent-p1-r7: not satisfied — event 28 answered withheld, and the probe expects absent session L2-campaign-absent-p1-r8: satisfied — event 29: answered absent session L2-campaign-absent-p1-r9: satisfied — event 30: answered absent
  • HW-PROBE-a-session-names-the-event-that-makes-a-document-accepted (sufficiency, expects answered) session L2-campaign-absent-p2-r1: satisfied — event 31: answered merge session L2-campaign-absent-p2-r10: not satisfied — event 32 answered stamp, and the probe expects merge session L2-campaign-absent-p2-r11: satisfied — event 33: answered merge session L2-campaign-absent-p2-r12: satisfied — event 34: answered merge session L2-campaign-absent-p2-r13: not satisfied — event 35 answered stamp, and the probe expects merge session L2-campaign-absent-p2-r14: satisfied — event 36: answered merge session L2-campaign-absent-p2-r15: satisfied — event 37: answered merge session L2-campaign-absent-p2-r16: satisfied — event 38: answered merge session L2-campaign-absent-p2-r17: satisfied — event 39: answered merge session L2-campaign-absent-p2-r18: satisfied — event 40: answered merge session L2-campaign-absent-p2-r19: satisfied — event 41: answered merge session L2-campaign-absent-p2-r2: satisfied — event 42: answered merge session L2-campaign-absent-p2-r20: satisfied — event 43: answered merge session L2-campaign-absent-p2-r21: satisfied — event 44: answered merge session L2-campaign-absent-p2-r22: satisfied — event 45: answered merge session L2-campaign-absent-p2-r23: satisfied — event 46: answered merge session L2-campaign-absent-p2-r24: satisfied — event 47: answered merge session L2-campaign-absent-p2-r25: satisfied — event 48: answered merge session L2-campaign-absent-p2-r26: not satisfied — event 49 answered stamp, and the probe expects merge session L2-campaign-absent-p2-r27: satisfied — event 50: answered merge session L2-campaign-absent-p2-r28: not satisfied — event 51 answered stamp, and the probe expects merge session L2-campaign-absent-p2-r29: satisfied — event 52: answered merge session L2-campaign-absent-p2-r3: satisfied — event 53: answered merge session L2-campaign-absent-p2-r30: satisfied — event 54: answered merge session L2-campaign-absent-p2-r4: satisfied — event 55: answered merge session L2-campaign-absent-p2-r5: satisfied — event 56: answered merge session L2-campaign-absent-p2-r6: not satisfied — event 57 answered stamp, and the probe expects merge session L2-campaign-absent-p2-r7: satisfied — event 58: answered merge session L2-campaign-absent-p2-r8: not satisfied — event 59 answered stamp, and the probe expects merge session L2-campaign-absent-p2-r9: satisfied — event 60: answered merge
  • HW-PROBE-a-session-names-the-status-a-settled-decision-carries-in-its-pull-request (sufficiency, expects answered) session L2-campaign-absent-p3-r1: not satisfied — event 61 answered current, which the probe does not declare session L2-campaign-absent-p3-r10: not satisfied — event 62 answered current, which the probe does not declare session L2-campaign-absent-p3-r11: not satisfied — event 63 answered current, which the probe does not declare session L2-campaign-absent-p3-r12: not satisfied — event 64 answered current, which the probe does not declare session L2-campaign-absent-p3-r13: not satisfied — event 65 answered current, which the probe does not declare session L2-campaign-absent-p3-r14: not satisfied — event 66 answered current, which the probe does not declare session L2-campaign-absent-p3-r15: not satisfied — event 67 answered current, which the probe does not declare session L2-campaign-absent-p3-r16: not satisfied — event 68 answered current, which the probe does not declare session L2-campaign-absent-p3-r17: not satisfied — event 69 answered current, which the probe does not declare session L2-campaign-absent-p3-r18: not satisfied — event 70 answered current, which the probe does not declare session L2-campaign-absent-p3-r19: not satisfied — event 71 answered current, which the probe does not declare session L2-campaign-absent-p3-r2: not satisfied — event 72 answered current, which the probe does not declare session L2-campaign-absent-p3-r20: not satisfied — event 73 answered current, which the probe does not declare session L2-campaign-absent-p3-r21: not satisfied — event 74 answered current, which the probe does not declare session L2-campaign-absent-p3-r22: not satisfied — event 75 answered current, which the probe does not declare session L2-campaign-absent-p3-r23: not satisfied — event 76 answered current, which the probe does not declare session L2-campaign-absent-p3-r24: not satisfied — event 77 answered current, which the probe does not declare session L2-campaign-absent-p3-r25: not satisfied — event 78 answered current, which the probe does not declare session L2-campaign-absent-p3-r26: not satisfied — event 79 answered current, which the probe does not declare session L2-campaign-absent-p3-r27: not satisfied — event 80 answered current, which the probe does not declare session L2-campaign-absent-p3-r28: not satisfied — event 81 answered current, which the probe does not declare session L2-campaign-absent-p3-r29: not satisfied — event 82 answered current, which the probe does not declare session L2-campaign-absent-p3-r3: not satisfied — event 83 answered current, which the probe does not declare session L2-campaign-absent-p3-r30: not satisfied — event 84 answered current, which the probe does not declare session L2-campaign-absent-p3-r4: not satisfied — event 85 answered current, which the probe does not declare session L2-campaign-absent-p3-r5: not satisfied — event 86 answered current, which the probe does not declare session L2-campaign-absent-p3-r6: not satisfied — event 87 answered current, which the probe does not declare session L2-campaign-absent-p3-r7: not satisfied — event 88 answered current, which the probe does not declare session L2-campaign-absent-p3-r8: not satisfied — event 89 answered current, which the probe does not declare session L2-campaign-absent-p3-r9: not satisfied — event 90 answered current, which the probe does not declare
  • HW-PROBE-a-session-records-an-unmeasured-claim-in-the-shape-this-corpus-checks (sufficiency, expects patched) session L2-campaign-absent-p4-r1: satisfied — event 91: section.required.missing reported nothing over docs/obligations/0223-c1-standing-context-is-charged-to-every-session-and-no-document-prices-it.md session L2-campaign-absent-p4-r10: satisfied — event 92: section.required.missing reported nothing over docs/obligations/0223-nothing-states-how-much-of-a-sessions-standing-instructions-cost-before-its-first-turn-of-work.md session L2-campaign-absent-p4-r11: satisfied — event 93: section.required.missing reported nothing over docs/obligations/0223-claude-md-costs-6-500-tokens-on-every-dispatch-and-nothing-turns-that-into-a-share-of-a-session-s-budget.md session L2-campaign-absent-p4-r12: satisfied — event 94: section.required.missing reported nothing over docs/obligations/0224-the-token-cost-of-c1-standing-context-has-no-stated-figure.md session L2-campaign-absent-p4-r13: satisfied — event 95: section.required.missing reported nothing over docs/obligations/0223-the-always-on-cost-of-standing-context-has-no-stated-size.md session L2-campaign-absent-p4-r14: satisfied — event 96: section.required.missing reported nothing over docs/obligations/0224-spec-5-says-standing-context-costs-a-session-and-no-document-states-how-much.md session L2-campaign-absent-p4-r15: satisfied — event 97: section.required.missing reported nothing over docs/obligations/0219-the-context-cost-of-claude-md-and-agents-md-before-a-session-s-first-turn-is-unmeasured.md session L2-campaign-absent-p4-r16: satisfied — event 98: section.required.missing reported nothing over docs/obligations/0224-a-sessions-standing-instructions-have-a-token-cost-and-nothing-here-has-ever-measured-one.md session L2-campaign-absent-p4-r17: satisfied — event 99: section.required.missing reported nothing over docs/obligations/0224-the-size-of-the-always-on-prompt-this-repository-pays-every-session-is-unmeasured.md session L2-campaign-absent-p4-r18: satisfied — event 100: section.required.missing reported nothing over docs/obligations/0224-nothing-states-how-much-of-a-sessions-budget-the-standing-context-spends-before-any-work-starts.md session L2-campaign-absent-p4-r19: satisfied — event 101: section.required.missing reported nothing over docs/obligations/0224-standing-instructions-cost-tokens-before-a-session-does-any-work.md session L2-campaign-absent-p4-r2: satisfied — event 102: section.required.missing reported nothing over docs/obligations/0223-the-standing-context-file-costs-its-size-on-every-session-and-nothing-states-what-that-size-is.md session L2-campaign-absent-p4-r20: satisfied — event 103: section.required.missing reported nothing over docs/obligations/0224-session-cost-cents-prices-every-arm-alike-and-no-comparison-reads-a-transcripts-realized-cost-by-arm.md session L2-campaign-absent-p4-r21: satisfied — event 104: section.required.missing reported nothing over docs/obligations/0223-nothing-states-how-much-of-a-session-s-budget-this-repository-s-standing-instructions-consume-before-work-starts.md session L2-campaign-absent-p4-r22: satisfied — event 105: section.required.missing reported nothing over docs/obligations/0223-each-standing-instruction-file-declares-its-own-ceiling-and-nothing-sums-what-a-session-pays-before-it-does-any-work.md session L2-campaign-absent-p4-r23: satisfied — event 106: section.required.missing reported nothing over docs/obligations/0223-nothing-states-how-much-of-a-session-s-context-budget-this-repository-s-standing-instructions-spend-before-work-starts.md session L2-campaign-absent-p4-r24: satisfied — event 107: section.required.missing reported nothing over docs/obligations/0224-a-declared-session-cost-cents-prices-a-whole-session-and-no-reading-separates-what-standing-instructions-spend-before-work-starts.md session L2-campaign-absent-p4-r25: not satisfied — the session produced no artifact, so the oracle had nothing to read session L2-campaign-absent-p4-r26: satisfied — event 109: section.required.missing reported nothing over docs/obligations/0224-spec-16-says-spec-5-prices-the-standing-instructions-cost-and-spec-5-states-the-argument-with-no-number.md session L2-campaign-absent-p4-r27: satisfied — event 110: section.required.missing reported nothing over docs/obligations/0224-nothing-states-how-many-tokens-of-a-sessions-standing-instructions.md session L2-campaign-absent-p4-r28: satisfied — event 111: section.required.missing reported nothing over docs/obligations/0224-nothing-states-how-much-of-a-sessions-budget-the-standing-instructions-of-this-repository-consume-before-any-work-starts.md session L2-campaign-absent-p4-r29: not satisfied — the session produced no artifact, so the oracle had nothing to read session L2-campaign-absent-p4-r3: satisfied — event 113: section.required.missing reported nothing over docs/obligations/0223-nothing-states-what-hw-run-policy-costs-a-session-before-its-work-begins.md session L2-campaign-absent-p4-r30: satisfied — event 114: section.required.missing reported nothing over docs/obligations/0223-no-record-measures-how-many-tokens-claude-md-costs-a-dispatch-before-any-work-starts.md session L2-campaign-absent-p4-r4: satisfied — event 115: section.required.missing reported nothing over docs/obligations/0224-nothing-states-how-much-of-a-sessions-budget-the-standing-instructions-of-this-repository-consume-before-any-work-starts.md session L2-campaign-absent-p4-r5: satisfied — event 116: section.required.missing reported nothing over docs/obligations/0224-claude-md-s-cost-lands-on-every-dispatch-and-no-document-states-its-size.md session L2-campaign-absent-p4-r6: not satisfied — the session produced no artifact, so the oracle had nothing to read session L2-campaign-absent-p4-r7: satisfied — event 118: section.required.missing reported nothing over docs/obligations/0224-spec-16-says-standing-context-costs-its-size-on-every-session-and-no-document-states-that-size.md session L2-campaign-absent-p4-r8: not satisfied — the session produced no artifact, so the oracle had nothing to read session L2-campaign-absent-p4-r9: not satisfied — the session produced no artifact, so the oracle had nothing to read

The rate, and the denominator it is over

76 of 120 graded sessions satisfied their expectation: 63.3%, in a 95% interval of 54.4% to 71.4%. The denominator is the graded sessions and never the selected probes. 0 sessions reached no verdict, and a session with no verdict is outside both halves of that fraction.

An interval that overlaps the previous run's is variance and one that does not is drift. This is one arm, so it estimates no effect: an efficacy claim is a comparison of two results, and the arm each one recorded is on it.

The comparisons this arm takes part in

What the governance changes. The treated arm is the campaign present arm and the control is the campaign absent arm, which removed the paths the campaign tier's ablation names and kept docs/.

  • treated, docs/probe-runs/campaign-of-2026-09-28-campaign-tier-present-arm-sufficiency.md: 71 of 120 graded sessions satisfied their expectation, 59.2%, in a 95% interval of 50.2% to 67.5%. 0 sessions refused by the session itself.
  • control, docs/probe-runs/campaign-of-2026-09-28-campaign-tier-absent-arm-sufficiency.md: 76 of 120 graded sessions satisfied their expectation, 63.3%, in a 95% interval of 54.4% to 71.4%. 0 sessions refused by the session itself.

The difference is -4.2 points, in a 95% Newcombe interval of -16.2 points to +8.1 points. The transcript of at least one arm is not current, so this page states no direction at the 5% level.

What the documents change. The treated arm is the campaign absent arm and the control is the documentation absent arm. Both removed the governance, and only the control removed docs/, so this is the effect of the documents alone (spec 5).

  • treated, docs/probe-runs/campaign-of-2026-09-28-campaign-tier-absent-arm-sufficiency.md: 76 of 120 graded sessions satisfied their expectation, 63.3%, in a 95% interval of 54.4% to 71.4%. 0 sessions refused by the session itself.
  • control, docs/probe-runs/campaign-of-2026-09-28-documentation-tier-absent-arm-sufficiency.md: 22 of 120 graded sessions satisfied their expectation, 18.3%, in a 95% interval of 12.4% to 26.2%. 0 sessions refused by the session itself.

The difference is +45.0 points, in a 95% Newcombe interval of +33.1 points to +55.0 points. The transcript of at least one arm is not current, so this page states no direction at the 5% level.