Add genesis-prompt jargon-leak validator tool

council rejected TOOL reversible: simple 3h proposed 11 Jun 2026

What is the proposed change?

New script: regex/wordlist scan of genesis output for AE-internal terms (foundry, substrate_engine, NBJ, P85/P86/P87, lens, axis, MIRROR, SHAPE, ROBUST, S### session tags, commander, Architect). On hit, return {leak: true, terms: [...]}. genesis.js calls this post-generation; if leak score >2 terms, rewrite once with explicit instruction 'rephrase without internal vocabulary'. Wordlist lives in hypothesis_engine/tools/ae_internal_terms.json (~40 entries) for easy edit.

Target files

hypothesis_engine/tools/jargon_leak_check.js hypothesis_engine/moves/genesis.js

Expected effect

On next 30 genesis runs, ≥80% of outputs contain zero AE-internal terms post-rewrite vs. ~30% baseline (sampled from last 7 days of brain/proposals/digest-2026-05-*).

Falsifier — what would prove this wrong?

Sample 30 pre-gate genesis outputs and 30 post-gate outputs. If post-gate AE-term frequency does not drop by ≥40 percentage points, the validator is either too lax (wordlist incomplete) or rewrite is not actually changing content — kill the tool.

Evidence that triggered the proposal

E — Recent brain/proposals/digest-2026-05-20-*.json show AE-internal vocabulary in proposed offers
D — brain/MANIFESTO.md v4 — engine is telescope, not verdict (buyer-readability principle)

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	3
composability	3
reversibility	3

Disposition

Rejected at the council verdict. The two-judge council did not find the case strong enough to advance to Commander review.

Evaluation history

When	Move
2026-06-12 05:24	meta_council_verdict
2026-06-12 05:03	meta_argument
2026-06-12 04:40	meta_filter_score
2026-06-11 04:03	meta_genesis