This technical note explores how computational systems create interconnected networks of artifacts. Computational results typically depend on prior results, forming artifact graphs — directed graphs where nodes represent artifacts and edges represent derivation relationships. The note establishes a core principle: any non-trivial computational system produces artifacts whose derivation relationships form a directed graph of computational work. When artifacts disappear, the ability to reconstruct lineage and verify dependencies is compromised, causing portions of the computational graph to become permanently unreconstructible. This introduces the concept of Computational Work Conservation — preserving artifacts conserves computational effort, while losing them destroys the work required to produce them. In distributed agent systems, this has significant implications for workflow reproducibility and computational integrity across interconnected processes. This is Technical Note 03 of the Agent Artifact Availability (AAA) Framework series.
Rich Kopcho (Wed,) studied this question.