Clusters are primary; nodes remain secondary.
KNOWLEDGE COVERAGE / V131
Knowledge Coverage & Gap Ledger
See what the atlas maps, what public evidence metadata resolves, and what still needs work.
Every ratio names its denominator. Every gap keeps its reason. No composite score turns metadata breadth into scientific quality, source validation, or production readiness.
- Knowledge clusters
- 45
- primary denominator
- Resolved experiment link
- 32
- of 45 clusters
- Experiment-link gaps
- 13
- explicitly queued
- Validated claim edges
- 0
- metadata is not proof
Experiment IDs must exist in the public research catalog.
Gap codes and queue bands state why work remains.
Metadata coverage never becomes a scientific score.
INSPECTION WORKSPACE
Forty-five clusters. No hidden ranking.
Filter the fixed V130-derived ledger, inspect each cluster's explicit dimensions, and export the same deterministic records as JSON or CSV.
Metadata linkage, not a reproduction claim.
The denominator is always the knowledge cluster.
Registered source metadata is not claim support.
Missing scholarly fields are never invented.
Four limits carried into every row
NO_VALIDATED_CLAIM_SOURCE_EDGES0Registered source metadata cannot be presented as claim support.
NO_COMPLETE_SCHOLARLY_CITATION_TUPLES0Citation exports remain verification-labeled starters.
CLUSTER_LEVEL_DISCOVERY_ONLY45Associations cannot be interpreted as direct coverage of every node.
METRIC_ONTOLOGY_DEFERREDV132V131 does not infer metric equivalence or conversion.
Queue bands, not a score
- 01Legacy priority10
- 02Unreviewed node1
- 03Reproducibility gap10
- 04Source only0
- 05Maintain24
Queue order is a transparent next-audit sequence, not a quality, importance, funding, or product-readiness score.
PUBLIC EVIDENCE BOUNDARY
Coverage is not correctness.
The ledger measures inspectable public metadata and declared review states. It does not inspect private source bodies or execute a solver.
- Cluster-level onlyA related route, source, simulation, or experiment is not direct evidence for every node.
- Linkage onlyA resolved experiment record does not prove successful independent reproduction.
- Freshness onlyA recorded review date is not a live endpoint or source-validity check.
- No dispositionNo causality, calibration, recipe, yield, signoff, or production decision follows.