Monorepo

A monorepo is a single version-controlled repository that contains multiple projects, packages, or services. The appeal is real: atomic commits across boundaries, shared tooling, a single source of truth for dependencies, and the ability to refactor across the entire codebase in one change.

The tradeoffs are also real. Build times grow. CI pipelines get complicated. Ownership boundaries blur. The tooling that made the repo manageable at ten packages becomes the constraint at a hundred. Every decision that felt local — how to name things, where to put shared code, how to scope tests — turns out to have consequences for everyone.

What makes monorepos work isn’t the structure itself but the discipline built around it: clear ownership, scoped configuration, fast incremental builds, and explicit contracts between packages. Without that, a monorepo is just a big repo.

Posts here cover the practical side — tooling choices, workspace configuration, how to scope AI tools like Claude to different parts of a monorepo, and the architectural decisions that become unavoidable as the codebase grows.