I've used GitHub Copilot for a while, but it feels like the tooling has moved way beyond autocomplete. Now there are agents, reusable code blocks, automated testing loops, and all kinds of new workflows. Is there a good place to learn how these pieces fit together in practice?

]]>https://community.openpreservation.org//topic/533/is-there-a-good-guide-to-modern-ai-assisted-development-workflowsRSS for NodeMon, 29 Jun 2026 09:06:12 GMTSun, 28 Jun 2026 21:15:02 GMT60The gap between what these tools can do and what most people are actually getting out of them is pretty wide right now, and it's mostly a documentation and workflow problem rather than a capability one. A lot of it clicks way faster when you see it laid out as a system rather than a list of features. Worth digging into Tetrees documentation here: https://tetrees.ai/?utm_source=upv . Goes through how agents, reusable blocks, and testing loops work together as an actual workflow rather than separate things you bolt on. Makes the whole thing a lot easier to wrap your head around in practice.

]]>https://community.openpreservation.org//post/630https://community.openpreservation.org//post/630Sun, 28 Jun 2026 21:46:30 GMTThe autocomplete era feels like a lifetime ago at this point. What's changed isn't just the features, it's the whole mental model of how you actually work with these tools. Agents that can run test loops, catch their own failures, and iterate without you babysitting every step are genuinely a different category of thing. The tricky part is that most people are still using them like fancy autocomplete and leaving a ton on the table because nobody sat down and explained how the pieces actually connect.

]]>https://community.openpreservation.org//post/629https://community.openpreservation.org//post/629Sun, 28 Jun 2026 21:43:15 GMT