The more I work with AI agents, the more I like the idea of keeping work where the work actually happens. If the code already lives in GitHub, why should tasks sit in Jira, comments hide in Google Docs, and extra context get scattered across Trello and ten random tabs? One repo, one workflow, one place where both people and agents can read, understand, and move things forward.
What I really like here is not just the convenience, but the methodology. Issues become tasks, pull requests become discussion threads, comments stay attached to the actual change, and the whole flow becomes much easier to follow.
This is especially useful for agents, because they work much better when the context sits next to the action instead of being fragmented across tools.
And honestly, you do not even need to be a full-time developer to use this thinking. 🤓 GitHub can be used for content, research, docs, internal operations, and other structured work too. It is probably still not the most comfortable interface for everything, fair enough, but the idea itself is very strong: keep context, decisions, and execution in one place, and suddenly collaboration becomes much simpler for humans and AI alike 🙂
P.S: Psst, did you notice, how I used NI Agent label hehe?
