Agents Help For Opensource Maintainance

I’ve spent more on AI coding agents in the past few weeks than any subscription in my life, in the order of thousands of rupees. For open source developers, they are a productivity game changer, especially if you maintain dozens of projects.

Agents force you to have great docs. You can tell the agent to read the docs or part of the docs to understand. You can also improve by telling them to write tutorials, or make them crystal clear. Clear docs suddenly pay off in ways you didn’t expect.

I maintain 25+ projects with thousands of downloads. Maintaining them used to drain my time. Agents let me scale, but even massive gains start to feel normal.

Humans still matter. Complex integrations and edge cases need manual testing. Writing docs, tests, and fitting in with a codebase still takes effort.

Agents are powerful, but they lock you into US SaaS. I can’t wait for local, open-source models to handle open source contributions.