PRs NOT Welcome: How Top AI Open Source Projects Are Managing Thousands of Contributors
· Source: Latent Space
GitHub introduced pull requests almost twenty years ago, and until now most projects kept them open so anyone could submit changes. In the AI space, high‑profile open‑source repositories such as Flue and tldraw are moving away from that practice and no longer accept external PRs. Maintainers say that a large portion of contributions come from generative tools and prefer that their own automated agents create and manage merge requests.
To handle the flood of issues and proposals, many teams are adopting what they call “software factories.” These factories consist of specialized agent groups: some reproduce bugs, others implement fixes, and a third type reviews changes before handing them to a human for final merge. Vercel exemplifies this model with its AI SDK, which had more than a thousand open issues and nearly eight hundred pull requests by the end of June. The company deployed several agents to manage different stages of the workflow, reducing manual effort and speeding up community response.
This trend signals an evolution in the governance of AI‑driven open‑source projects, where automation scales review and maintenance without compromising quality. Understanding how these software factories operate is essential for participants in AI ecosystems, as it shapes how new ideas are integrated and project stability is maintained.
Read the original article on Latent Space
This summary is an informational synthesis produced by dataqbs.com. All rights to the original content belong to its author and the cited media outlet. We act solely as curators of technology news and claim no authorship.