knowledge-centric-ai

Reference Traces

Reference traces serialize governed executions within a stated assessment scope. Each trace connects Registry state, intent, resolution, assembly, execution evidence, governance, and resulting institutional change through stable identifiers.

Where relevant, traces also identify the overlapping semantic, procedural, and episodic roles played by governed knowledge. These labels describe functional roles in organizational memory; they do not introduce additional principal artifact types or prescribe separate storage layers.

Traces preserve the double role of a CEE: it consumes the governed memory selected into its assembly and may produce situated candidate knowledge. The trace must retain execution and producer identity while keeping production separate from institutional adjudication.

Between materialization and evidence submission, the CEE may run autonomously inside its bounded mandate. A conforming trace records the mandate, the exact materialization used, and the terminal or contract-defined checkpoint, but does not require step-wise Registry interaction or wholesale working-state capture.

Available traces

Reusable, consumer-independent object representations are catalogued separately under reference-objects/.