Matt Legrand
38 posts
Brief thoughts on @augmentcode : > Augment's output is much more accurate than Cursor & Cline/Kilo/etc. > Aug's Claude 4 is great for building 90% of the project > Aug's GPT5 is superb at spotting the finer details of that last difficult 10% > Enhanced prompt is MAGIC ✨
Claude Sonnet 4 now supports 1 million tokens of context on the Anthropic API—a 5x increase. Process over 75,000 lines of code or hundreds of documents in a single request.
Launch Week Day 1: Tasklist is here Eagle-eyed power users have already been putting Tasklist to work. Now, we’re showing you how to get the most out of it. With Tasklist, we’re making it way easier to take a big, vague task—and break it down step by step. So you can actually see what your Agent’s doing, monitor progress, and make changes as you go. It works the same way the most productive engineers do: - Define the ideal outcome - Break it into steps - Verify the plan makes sense That breakdown gives you confidence. Tasklist builds that into how the Agent works.
Codex sucks for phased multi-agent approaches. It always ends up in PR merging issues. It needs meticulous planning by the user. The most autonomous planning and execution currently comes from the remote agents of Augment Code. Factory droids come close but are slower and much more expensive. @augmentcode is awesome.
