AI Coding Agents Are Breaking Big Codebases — Dan Adler, Sourcegraph
- Source
- youtube.com
- Author
- AI Engineer
- Date
Context
A software migration across hundreds of repositories requires finding every affected project, adapting the edit and tracking review. Code intelligence company Sourcegraph describes Agentic Batch Changes as a way to coordinate that work from one request. Its searches the indexed codebase, plans the change, tries it in one repository and then expands the rollout.
The company says the system uses scripts for repetitive edits and coding agents where judgment is needed. It responds to failed automated checks and tracks proposed changes until they are ready for human review and merging. The useful capability is coordinated maintenance across projects whose setups differ, including visibility into which changes have actually landed. Read the workflow.
- Sourcegraph CEO Dan Adler argues the wave of agent-written code is making large, decades-old codebases decay through duplicated code, drifting standards, brittle dependencies and new vulnerabilities.
- His central claim is that at enterprise scale the bottleneck is , not model quality: agents understand code by searching, so code spread across tens of thousands of repositories stays invisible to them.
- Adler frames large-scale code visibility as infrastructure, and presents Agentic Batch Changes as a way to make and audit one change across thousands of repos from a single prompt, with a Mercari case study.
- AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
- context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
Checking sign-in…
Loading comments…






