Rendered from docs/probe-results/campaign-of-2026-10-03-campaign-tier-no-claude-md-arm-navigability.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-2026-10-03-campaign-tier-no-claude-md-arm-navigability.md

A probe result is a function of committed inputs and of nothing else: the transcript at docs/probe-runs/campaign-of-2026-10-03-campaign-tier-no-claude-md-arm-navigability.md and the state it stands at, the expectations the probes it names declare, the version of the grader that evaluated them, and for a compared arm the result of the other arm. Fetch those and this file comes back.

The run this transcript recorded

A campaign run in the no-claude-md arm, on claude-sonnet-5 at 2026-10-03. served version claude-sonnet-5, tree sha256:81d5f62acb289022175d192df0a61faf5f71eb4b79e8dd1f4efae4532ebdd916, selection sha256:77ce7e13166fd99f67da4bb4ddd90eaa6076dedfb335bc54563f94dce33b0238, read set sha256:18ae82cb33e60fa707b36ccc3090f962d3b82ac1a179359fdaecfc1a30e6be7b, seed 0, harness 0.5.0. realized cost $33.35, which the adaptive layer reads as the cost of its own instrument.

150 events over 5 probes, in 150 sessions and 918 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. This page names the lock, the selection and the read set the transcript recorded, and it compares none of them with the tree in front of a reader. headwater generate and headwater probe stale name what moved since the recording, and neither one fails for 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 sha256:77ce7e13166fd99f67da4bb4ddd90eaa6076dedfb335bc54563f94dce33b0238, and the probes graded below are the probes it was planned over, as this corpus declares them now. A probe added to this corpus after the recording 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. This file names it as provenance and compares it with nothing, and the same holds for the lock and the selection: an edit to a document, a new probe or a moved lock leaves these bytes alone. headwater generate names each result whose recorded lock, read set or selection moved, and headwater probe stale takes the digest and reports which recorded results a change voided. Neither one fails for it.

Graded by grader 0.5.0. A campaign run in the no-claude-md arm, on claude-sonnet-5 at served version claude-sonnet-5.

The verdicts

  • HW-PROBE-a-session-answers-from-the-register-without-opening-the-question-it-replaced (navigability, expects not_opened) session L19-campaign-no-claude-md-p2-r1: satisfied — 6 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r10: satisfied — 6 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r11: satisfied — 7 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r12: satisfied — 7 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r13: satisfied — 3 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r14: satisfied — 6 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r15: satisfied — 4 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r16: satisfied — 5 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r17: satisfied — 7 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r18: satisfied — 9 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r19: satisfied — 5 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r2: satisfied — 4 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r20: satisfied — 5 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r21: satisfied — 5 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r22: satisfied — 6 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r23: satisfied — 3 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r24: satisfied — 5 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r25: satisfied — 3 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r26: satisfied — 8 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r27: satisfied — 3 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r28: satisfied — 5 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r29: satisfied — 3 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r3: satisfied — 6 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r30: satisfied — 4 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r4: satisfied — 3 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r5: satisfied — 3 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r6: satisfied — 5 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r7: satisfied — 9 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r8: satisfied — 6 recorded calls, and none named any of the 1 document session L19-campaign-no-claude-md-p2-r9: satisfied — 5 recorded calls, and none named any of the 1 document
  • HW-PROBE-a-session-says-how-a-governs-line-reaches-a-document (navigability, expects answered) session L19-campaign-no-claude-md-p3-r1: not satisfied — event 61 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r10: satisfied — event 62: answered command session L19-campaign-no-claude-md-p3-r11: satisfied — event 63: answered command session L19-campaign-no-claude-md-p3-r12: satisfied — event 64: answered command session L19-campaign-no-claude-md-p3-r13: not satisfied — event 65 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r14: satisfied — event 66: answered command session L19-campaign-no-claude-md-p3-r15: satisfied — event 67: answered command session L19-campaign-no-claude-md-p3-r16: satisfied — event 68: answered command session L19-campaign-no-claude-md-p3-r17: not satisfied — event 69 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r18: not satisfied — event 70 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r19: not satisfied — event 71 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r2: satisfied — event 72: answered command session L19-campaign-no-claude-md-p3-r20: satisfied — event 73: answered command session L19-campaign-no-claude-md-p3-r21: not satisfied — event 74 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r22: not satisfied — event 75 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r23: not satisfied — event 76 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r24: not satisfied — event 77 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r25: satisfied — event 78: answered command session L19-campaign-no-claude-md-p3-r26: satisfied — event 79: answered command session L19-campaign-no-claude-md-p3-r27: not satisfied — event 80 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r28: satisfied — event 81: answered command session L19-campaign-no-claude-md-p3-r29: satisfied — event 82: answered command session L19-campaign-no-claude-md-p3-r3: not satisfied — event 83 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r30: not satisfied — event 84 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r4: not satisfied — event 85 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r5: not satisfied — event 86 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r6: not satisfied — event 87 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r7: not satisfied — event 88 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r8: not satisfied — event 89 answered hand, and the probe expects command session L19-campaign-no-claude-md-p3-r9: not satisfied — event 90 answered hand, and the probe expects command
  • HW-PROBE-a-session-says-where-a-figure-on-the-site-is-written (navigability, expects answered) session L19-campaign-no-claude-md-p4-r1: satisfied — event 91: answered published session L19-campaign-no-claude-md-p4-r10: satisfied — event 92: answered published session L19-campaign-no-claude-md-p4-r11: satisfied — event 93: answered published session L19-campaign-no-claude-md-p4-r12: satisfied — event 94: answered published session L19-campaign-no-claude-md-p4-r13: satisfied — event 95: answered published session L19-campaign-no-claude-md-p4-r14: satisfied — event 96: answered published session L19-campaign-no-claude-md-p4-r15: satisfied — event 97: answered published session L19-campaign-no-claude-md-p4-r16: satisfied — event 98: answered published session L19-campaign-no-claude-md-p4-r17: satisfied — event 99: answered published session L19-campaign-no-claude-md-p4-r18: satisfied — event 100: answered published session L19-campaign-no-claude-md-p4-r19: satisfied — event 101: answered published session L19-campaign-no-claude-md-p4-r2: satisfied — event 102: answered published session L19-campaign-no-claude-md-p4-r20: satisfied — event 103: answered published session L19-campaign-no-claude-md-p4-r21: satisfied — event 104: answered published session L19-campaign-no-claude-md-p4-r22: satisfied — event 105: answered published session L19-campaign-no-claude-md-p4-r23: satisfied — event 106: answered published session L19-campaign-no-claude-md-p4-r24: satisfied — event 107: answered published session L19-campaign-no-claude-md-p4-r25: satisfied — event 108: answered published session L19-campaign-no-claude-md-p4-r26: satisfied — event 109: answered published session L19-campaign-no-claude-md-p4-r27: satisfied — event 110: answered published session L19-campaign-no-claude-md-p4-r28: satisfied — event 111: answered published session L19-campaign-no-claude-md-p4-r29: satisfied — event 112: answered published session L19-campaign-no-claude-md-p4-r3: satisfied — event 113: answered published session L19-campaign-no-claude-md-p4-r30: satisfied — event 114: answered published session L19-campaign-no-claude-md-p4-r4: satisfied — event 115: answered published session L19-campaign-no-claude-md-p4-r5: satisfied — event 116: answered published session L19-campaign-no-claude-md-p4-r6: satisfied — event 117: answered published session L19-campaign-no-claude-md-p4-r7: satisfied — event 118: answered published session L19-campaign-no-claude-md-p4-r8: satisfied — event 119: answered published session L19-campaign-no-claude-md-p4-r9: satisfied — event 120: answered published
  • HW-PROBE-an-agent-reaches-the-adjudication-from-the-document-that-lost-it (navigability, expects opened) session L19-campaign-no-claude-md-p1-r1: satisfied — event 1, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r1/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r10: satisfied — event 2, call 2: Read docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r11: satisfied — event 3, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r11/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r12: satisfied — event 4, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r12/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r13: satisfied — event 5, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r13/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r14: satisfied — event 6, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r14/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r15: satisfied — event 7, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r15/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r16: satisfied — event 8, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r16/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r17: satisfied — event 9, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r17/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r18: satisfied — event 10, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r18/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r19: satisfied — event 11, call 3: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r19/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r2: satisfied — event 12, call 4: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r2/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r20: satisfied — event 13, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r20/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r21: satisfied — event 14, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r21/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r22: satisfied — event 15, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r22/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r23: satisfied — event 16, call 4: Read docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r24: satisfied — event 17, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r24/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r25: satisfied — event 18, call 5: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r25/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r26: satisfied — event 19, call 2: Read docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r27: satisfied — event 20, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r27/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r28: satisfied — event 21, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r28/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r29: satisfied — event 22, call 3: Read docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r3: satisfied — event 23, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r3/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r30: satisfied — event 24, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r30/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r4: satisfied — event 25, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r4/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r5: satisfied — event 26, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r5/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r6: satisfied — event 27, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r6/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r7: satisfied — event 28, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r7/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r8: satisfied — event 29, call 5: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r8/docs/decisions/0008-probe-cost-and-cadence.md session L19-campaign-no-claude-md-p1-r9: satisfied — event 30, call 2: Read /var/tmp/hw-1659/a/ws/L19-campaign-no-claude-md-p1-r9/docs/decisions/0008-probe-cost-and-cadence.md
  • HW-PROBE-what-a-session-writes-points-back-at-the-ruling-it-rests-on (navigability, expects cited) session L19-campaign-no-claude-md-p5-r1: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r10: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r11: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r12: satisfied — event 124: engine/crates/census/src/census.rs cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r13: satisfied — event 125: docs/explanations/why-a-corpus-wide-count-is-derived-not-stored.md cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r14: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r15: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r16: satisfied — event 128: docs/explanations/why-a-corpus-wide-count-is-computed-when-it-is-read.md cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r17: satisfied — event 129: docs/explanations/why-a-corpus-wide-count-is-computed-on-read-not-kept-in-a-committed-file.md cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r18: satisfied — event 130: docs/explanations/why-a-corpus-wide-count-is-derived-at-read-time-never-stored.md cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r19: satisfied — event 131: docs/explanations/why-a-corpus-wide-count-is-computed-at-read-time-and-not-stored.md cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r2: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r20: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r21: satisfied — event 134: docs/explanations/why-a-corpus-wide-count-is-computed-when-read-not-committed.md cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r22: satisfied — event 135: docs/explanations/a-corpus-wide-count-is-computed-when-the-corpus-is-read.md cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r23: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r24: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r25: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r26: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r27: satisfied — event 140: docs/evaluations/why-corpus-counts-are-derived-not-stored.md cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r28: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r29: satisfied — event 142: docs/explanations/why-a-count-over-the-whole-corpus-is-computed-when-read-not-committed.md cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r3: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r30: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r4: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r5: satisfied — event 146: docs/explanations/why-a-corpus-wide-count-is-computed-when-read-not-committed.md cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r6: satisfied — event 147: engine/crates/census/src/census.rs cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r7: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier session L19-campaign-no-claude-md-p5-r8: satisfied — event 149: engine/crates/census/src/census.rs cites HW-DR-0049 session L19-campaign-no-claude-md-p5-r9: not satisfied — 0 produced artifacts, and none cites any of the 1 identifier

The rate, and the denominator it is over

115 of 150 graded sessions satisfied their expectation: 76.7%, in a 95% interval of 69.3% to 82.7%. 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 CLAUDE.md changes. The treated arm is the campaign present arm and the control is the campaign no-claude-md arm, which removed the paths the no-claude-md component's delta names. The documents are in both arms (spec 5).

  • treated, docs/probe-runs/campaign-of-2026-10-03-campaign-tier-present-arm-navigability.md: 95 of 150 graded sessions satisfied their expectation, 63.3%, in a 95% interval of 55.4% to 70.6%. 0 sessions refused by the session itself.
  • control, docs/probe-runs/campaign-of-2026-10-03-campaign-tier-no-claude-md-arm-navigability.md: 115 of 150 graded sessions satisfied their expectation, 76.7%, in a 95% interval of 69.3% to 82.7%. 0 sessions refused by the session itself.

The difference is -13.3 points, in a 95% Newcombe interval of -23.3 points to -3.0 points. The interval is below zero, so the treated arm satisfied less often at the 5% level.