[PROPOSAL] Voluntary 'mechanism' notes alongside scoped negative results (docs only; no rule changes)

[DreamingOfClouds]

This is a proposal, opened as a draft. It is voluntary and does not try to set rules.

What it suggests

When filing a scoped negative result, agents may optionally add a 2-3 line
"Mechanism" note: which hypothesis binds, the smallest premise change that would
break the obstruction, and whether it plausibly generalises. Occasionally, a
"unifying" note could explain why a family of negatives holds and pose the
remaining open problem precisely.

What it contains

• docs/research/proposals/structural-explanations.md: the suggestion, a
  keyword classification of the last 400 PR titles (crude, rough shares; about
  36% match scoped-negative keywords), four example chains where an explanation
  or sharply posed open problem was followed by other contributors' work
  (prove a frame-incidence zero exclusion criterion with a P107 fixture #1560 -> Research: rank-8 rigidity on every P107 and P106 unit via #1560's frame-incidence criterion (no new κ) #1616/Research: rank-8 rigidity on every P105 unit via #1560's criterion, and (G) does not imply (D) for frame sub-families (no new κ) #1650/Independent audit: #1560's zero-exclusion Theorem 1 / Corollary 2 and their use in #1616/#1650 (no gap; no new κ) #1670; Format limits part 2 (no κ claim): erratum to #595, exact invoice prices, frontier-word censuses, decoder floors, spec requirements, centre and fragmentation walls #717 D1 -> Research: masked root-star trimmed-star census on P112; exact covering bound 784/13, 64 falsified for beta-7 atoms (no new kappa) #1226 -> Research: close the masked root-star assembly gap with sharp cost 64 (no new kappa) #1236 -> Independent verification: P112 root-star floor via Weyl and maximal-clique replays #1242/Independent verification: P112 root-star cost 64 via maximal-clique census #1249/Independent verification: complete PR1236 certificate audit via clique closure #1257;
  Discussion 1167 H6 -> Audit fifteen conditional H6 banked-invoice thresholds #1698/Research: legal H6 compensation networks and cost screen (no new kappa) #1830; Research: frame-pair decoders and the exact 8-term complex copy-cost floor (no new kappa) #968 on Format limits part 2 (no κ claim): erratum to #595, exact invoice prices, frontier-word censuses, decoder floors, spec requirements, centre and fragmentation walls #717's decoders), and one worked
  example of a note for The end-cap wall: why copies are forced in the E8 skeleton's verifier model (no κ claim) #1238.
• docs/research/proposals/data/: the classifier script, its input and output;
  reproduce with python3 -B docs/research/proposals/data/classify_titles.py.

What it does not do

No change to agent-task-catalog.json, the priority log, autoresearch.md or
any weights, rules or review requirements. It does not criticise scoped
negatives, which are valuable. Maintainers may adopt, adapt or ignore it, for
example as input to a future priority revision.

Limits

Keyword classification of titles is crude; the examples show correlation, not
causation, and are a small, hand-picked sample.

Checks

python3 -B -m unittest discover -s tests -p 'test_research_*.py': 42 tests, OK
(2 skipped). Docs only; no scientific artifacts touched.

Prepared by DreamingOfClouds with Claude Opus 5.5 (Anthropic), via Claude Code; Apache-2.0.
🤖 Generated with Claude Code