Add 6th filter `named_post_mortem_data_source` to filter_score.js FILTERS

filter rejected AXIS reversible: simple 5h proposed 5 Jun 2026

What is the proposed change?

Insert as 6th element in the FILTERS array (line 13-19): `{ key: 'named_post_mortem_data_source', question: 'Does the hypothesis name a specific source of proprietary post-incident artefacts (failure trajectories, agent traces, IC memos, support escalations, change orders) only this buyer can produce, with the surface where they accumulate?' }`. Extend HIGH and LOW system prompts with a rubric paragraph: Score 3 = names a concrete artefact + named storage surface (e.g. 'LangSmith trace exports from buyer's prod agents'); Score 2 = artefact named, surface inferred; Score 1 = generic 'logs' or 'outputs' with no artefact specificity; Score 0 = the moat is described as 'pattern library' or 'curation' with no source-of-truth named. Update JSON output schema in both prompts to include the new key. Composite formula in line 211 unchanged (still uses summed midpoint over FILTERS).

Target files

hypothesis_engine/moves/filter_score.js

Expected effect

Among graduated candidates whose data_moat scored 2-3, those whose moat language is hand-waved (no named trajectory source) score 0-1 on the new axis, producing a 2-3 point composite spread that re-ranks vague-moat hypotheses below specifically-sourced ones. Forecaster-v1-class hypotheses (specific trajectory sources) score 2-3; oracle-archetype descendants where the 'data' is general financial commentary score 0-1.

Falsifier — what would prove this wrong?

Re-score the 43 graduated candidates currently in `engine.db` against the new axis. If the new axis is uncorrelated (|r| < 0.25) with the existing data_moat axis, it is redundant. If it is correlated >0.85 with data_moat, it provides no new information and should be removed.

Evidence that triggered the proposal

E — brain/proposals/digest-2026-06-04-007.json — Opus review: 'proprietary post-mortem trajectories from failed agent runs are exactly the data hyperscalers can't replicate cross-customer — that's a sharper wedge description than Agent Output Quality Benchmarking'
B — S183 close + MANIFESTO v4: forecaster-v1 wedge pivot is precisely about naming a specific resolvable-evidence source; current data_moat axis does not enforce naming

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
2026-06-12 04:15	meta_filter_score
2026-06-05 04:04	meta_genesis