Eli Bendersky on starting new projects with LLM agents, based on building a new Go project from scratch

grounded workflow advice from someone shipping a maintainable codebase, not just prototyping; especially good on small-CL discipline and the distinction between cheap code generation and expensive long-term maintenance.

Original source

What it is: Eli Bendersky on starting new projects with LLM agents, based on building a new Go project from scratch.

Gist: argues agent-heavy development works best when humans keep tight control over design, review, and commit boundaries: start with repo-committed design notes, keep CLs small and reviewable, use strong external tests, and avoid vibe-coding for projects you intend to maintain. He also makes the case that Go is especially agent-friendly because human time shifts from writing to reading.

Newsletter angle: “agent coding turns programming into a reading-heavy discipline” or “small CLs and strong tests are what make agent-built projects maintainable.”

Note: fetched article directly and extracted successfully.