Scaling Git for the agentic software factory

GitLab argues that AI agents are increasing software-development activity beyond the capacity of traditional Git-based workflows, citing Anthropic’s growth from roughly 2,500 to more than 30,000 daily monorepo merges and its creation of the specialized Scion system. GitLab’s next-generation source code management system, currently in private beta, aims to retain Git compatibility while using horizontally scalable stateless compute, shared object storage, and agent-native interfaces that avoid repeated full repository clones. Early internal tests reported up to 50 times faster workloads, up to twice fewer tokens, and up to 1,000 times less network traffic under specific conditions, though GitLab says further testing is needed under sustained concurrent use. The company positions SCM as one component of a broader “governed software factory” that also addresses CI/CD capacity, security, context, orchestration, and accountability for agent actions through products including GitLab Orbit and the Duo Agent Platform. GitLab says the architecture can operate on GitLab.com or customer-controlled S3-compatible and air-gapped infrastructure, preserving data sovereignty while supporting large-scale agent workloads, and it plans to continue collaborating with the open Git community.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.