Blog
- The Bitter Lesson of LLM Extensions11/24/2025
From ChatGPT Plugins to Agent Skills, a look at how we've been trying (and failing) to extend LLMs for the last three years.
- Cursor's Composer Model Surprised Me11/13/2025
After months away from Cursor, Composer 1 has me reconsidering my daily driver. Speed beats raw intelligence for most coding tasks.
- The Rise of Codex9/6/2025
How GPT-5's Codex went from almost unused to 28% of coding agent usage in just one month.