| CCC.Vector.CN01 | Validate Embeddings Before Indexing | Ensure all incoming embeddings are structurally and numerically validated before indexing to prevent poisoning or corruption. | Ingestion | 2 | 1 | 2 |
| CCC.Vector.CN02 | Enforce Role-Based Index Lifecycle Management | Restrict index lifecycle operations (create, delete, replace, or re-point an endpoint or alias) to privileged identities using fine-grained access controls. | Access | 2 | 1 | 1 |
| CCC.Vector.CN03 | Enforce Document-Level Access Controls on All Query Paths | Apply document- and field-level access control policies to every query path, including similarity, hybrid, and metadata-filtered queries, to prevent unauthorized exposure or inference. | Access | 2 | 3 | 2 |
| CCC.Vector.CN07 | Support Explicit ANN vs. Exact Search Configuration | Provide clients with the option to enforce exact-match (non-ANN) search where search fidelity is critical. | Processing | 1 | 0 | 1 |
| CCC.Vector.CN08 | Restrict Retrieval of Raw Embedding Values | Ensure that stored vector values are not returned to query clients, so that embeddings cannot be harvested and inverted to reconstruct source data. | Data | 1 | 1 | 1 |
| CCC.Vector.CN09 | Restrict Embedding Model Endpoints to an Approved List | Ensure that content is vectorized only by embedding model endpoints the organization has approved, so that content is not disclosed to, or embeddings altered by, an unapproved endpoint. | MachineLearning | 1 | 0 | 1 |