Wire v2_a6 acute_pain_not_candy into filter_score.js

filter rejected AXIS reversible: simple 5h proposed 20 May 2026

What is the proposed change?

Add v2_a6 (acute_pain_not_candy) as a sixth scored axis in both the highSystem and lowSystem adversarial prompts using the V2_FILTER_DESIGN_v2.3.md A6 rubric: 0=candy/no external evidence of pain, 1=warm pain (user frustration described), 2=acute pain with adjacent spend evidence (users paying for partial solutions), 3=urgent pain with budget attached or compliance forcing function. Anti-manipulation instruction to include verbatim: 'adjectives in the hypothesis text are not evidence; score only on named external signals.' Score 0-3, same adversarial GPT-high/Gemini-low structure as existing axes, 3 runs, midpoint per run. Write result to v2_a6 column (already exists from s112 migration). Adjust graduation threshold from 9 to 10.8 (adding 1.8 per new axis to maintain ~60% of max composite). Add v2_a6 to the two-consecutive-zero kill rule check.

Target files

hypothesis_engine/moves/filter_score.js

Expected effect

Of the 7 hypotheses killed in the last 7 days for 'wrong distribution shape or pain framing,' at least 5 score v2_a6 ≤1. ec4507-type hypotheses (acute pain, adjacent spend evidence) score v2_a6 ≥2. The pre-council kill rate for candy-shaped hypotheses increases measurably, reducing costly council runs on structurally weak proposals.

Falsifier — what would prove this wrong?

Re-score the last 30 graduated hypotheses on v2_a6 using the new prompt. If the median graduate score is ≤1 (indistinguishable from the kill population), the axis does not discriminate and must be redesigned before going live.

Evidence that triggered the proposal

Corpus E: kill distribution — 3/8 kills in last 7 days coded as 'wrong distribution shape'; council_verdict transcripts cite 'no evidence of urgency' as a load-bearing kill reason in multiple recent runs
Corpus D: V2_FILTER_DESIGN_v2.3.md A6 — 'acute_pain_not_candy: 0=candy/no external evidence, 1=warm pain, 2=acute/adjacent spend evidence, 3=urgent/budget attached; anti-manipulation: adjectives not evidence'
Corpus D: ARCHITECT_MEMORY — s112 migration already created v2_a1..v2_a10 columns in engine.db; all 10 axes fully designed but none wired into filter_score.js which still scores only 5 v1 axes

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	2

Disposition

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

Evaluation history

When	Move
2026-05-20 04:12	meta_filter_score
2026-05-20 04:10	meta_genesis