Repository navigation
Account for two source-backed mixed-scope records in connectivity census #1344
Description
Legitimate Scope Corrections Reach A Corpus Guard
At reviewed batch21 head b4fd7d01144bf9815c3c5b5ce8c1c7b628a0a018, the full both-root connectivity survey reports 120 mixed-scope records, exactly the current upper bound in tests/test_community_level_connectivity_credit.py.
The next source-backed repairs should represent JN4-GD17 product output and staged-butanol product output as community outcomes while retaining their genuine organism-level mechanisms. This is expected to add exactly two mixed-scope records, with no new graph nodes or arrows.
Verify the actual before/after census and account for those two records. If confirmed, change only the mixed-scope ceiling from 120 to 122 and document why. Preserve all other census bounds, connectivity behavior, participant exact-set checks and source-specific regression coverage. Do not misclassify the scientific records to preserve an obsolete count.
Dedup: #1203 is closed and concerns a different ceiling (380 community-level records), not this mixed-scope bound. #312 and #524 document the underlying audit behavior; no change to that behavior is proposed. All 678 issue titles/bodies and ignored/hidden local reports/tests/history/research were searched; comments not exhaustive.
Activity
realmarcin commentedon Oct 6, 2026
Addressed in #1346 at
d71b3800c819fff8b4c5a32bb99d0f68e5ca85c3. Exact-head source-based self-adversarial review: #1346 (review). All 30 local gates and actual post-commit docs-current passed; 270 passed in 237.82s (0:03:57); 87 MATCH snippets, zero other classes; 3 passed in 2.14s; 55 unrelated raw blocks and seven caches byte-preserved. See committed batch22 decision/adversarial/validation records for assay bounds and full node/arrow accounting. This issue remains open until protected main merge; #1343 is separate non-graph work.