Rethinking Source Control for Agentic Software Engineering: Architectural Transformation, Data Models, and Scalability Constraints in AI-Driven Development

This paper reconsiders the architectural requirements of source control systems when they serve parallel, agentic workloads rather than human-paced collaborative development. The central research question is reframed: not whether a specific system (e.g., Scion) is "better than" Git, but which architectural changes become necessary — versus optional design choices — when source control must support many concurrent autonomous agents operating on the same repository. The analysis is organized around five axes: workload characteristics, repository architecture, data architecture, execution infrastructure, and governance/security. The paper proposes a three-layer design: - Layer A (Architectural Analysis): What source control must do differently under agentic workloads. - Layer B (Structural Analysis): How the five axes interact and constrain each other. - Layer C (Empirical Evaluation): A proposed, unexecuted protocol for future benchmarking. Three critical corrections are built into the analysis as binding constraints: 1. Scion did not eliminate merge conflicts; it shifted them to post-merge/pre-release testing. 2. Scion is not unconditionally lock-free; changeset creation avoids full coordination, but ref updates still use CAS. 3. Derived data is treated as an architectural decision, not a storage optimization. Two open research questions are posed: 1. What is the latency cost of on-demand derived-data computation? 2. What is the scope of "functional decentralization" in practice? All operational figures attributed to Anthropic (30,000+ merges/day, >99% cache hit rate, ~200× infrastructure reduction, ~100× checkout speedup, 25% CI improvement) are explicitly flagged throughout as Anthropic's own self-reported figures, not independently verified. References include: Potvin & Levenberg (2016) on Piper, Corbett et al. (2013) on Spanner, Sapling/Mononoke documentation, and the Anthropic source-control article. Keywords: source control; version control; agentic workloads; distributed systems; repository architecture; merge conflicts; functional decentralization; Scion; Git; Piper; Spanner. Author: Samir Baladi (Ronin Institute / Rite of Renaissance, ORCID: 0009-0003-8903-0029) DOI: https://doi.org/10.5281/zenodo.23934465 License: CC-BY-4.0 (manuscript).