Add v2_a11 audit_shape_penalty axis to filter_score

filter rejected AXIS reversible: simple 4h proposed 17 Jun 2026

What is the proposed change?

Add axis v2_a11 'audit_shape_penalty' joining v2_a1..v2_a10 in the composite. Score 0-3 INVERTED: 3 = no audit/compliance/dashboard surface, 0 = pure audit/observability play. Detection prompt scores three signals from the candidate's frame field: (a) does the verb lean 'audit/score/grade/inspect/review', (b) is the buyer paying for a verdict vs an action, (c) is the artifact a report/dashboard vs a workflow output. Sum -> banded score. Insert after the v2_a10 block in filter_score.js and add to composite sum. Composite threshold stays at 17/30 (now /33); document the recalibration in V2_FILTER_DESIGN.

Target files

hypothesis_engine/moves/filter_score.js

Expected effect

On the 4 Commander-overridden KILL cases (all audit-shaped per Corpus E override log), v2_a11 scores 0-1 producing composite drop of 2-3 points vs prior runs. At least 3 of the 4 fall below 17/33 composite and would have been auto-killed without override.

Falsifier — what would prove this wrong?

Replay the 4 overridden candidates and the last 40 KEEP-graduated candidates. If v2_a11 score does not show >=2 point mean separation between Commander-killed-audit set and graduated set, the axis is not discriminating and should be removed.

Evidence that triggered the proposal

E — kill_reason_distribution: 4 commander_override_kill events all with frame=audit/compliance
D — brain/V2_FILTER_DESIGN_v2.3.md — 10-axis set with no audit-shape penalty

Proposer self-score

The proposer scored its own draft on these axes (0-3 each) before submitting.

Axis	Score
specificity	3
falsifier	3
solo feasible	3
blast radius	2
composability	3
reversibility	3

Disposition

Rejected by filter_score. The proposal did not meet the bar for specificity, falsifiability, or solo-feasibility.

Evaluation history

When	Move
2026-06-17 04:03	meta_filter_score
2026-06-17 04:03	meta_genesis