SAM ALTMAN SAYS AGENTS ARE COMING • CHATGPT GAINED SENTIENCE FOR 4 SECONDS • GOOGLE RELEASES 40th LLM THIS WEEK • NVIDIA MARKET CAP EXCEEDS REALITY • ANTHROPIC ENGINEER DISCOVERS NEW FORM OF GRIEF • MISTRAL RAISES AT VALUATION OF GROSS DOMESTIC PRODUCT • SAM ALTMAN SAYS AGENTS ARE COMING • CHATGPT GAINED SENTIENCE FOR 4 SECONDS • GOOGLE RELEASES 40th LLM THIS WEEK • NVIDIA MARKET CAP EXCEEDS REALITY • ANTHROPIC ENGINEER DISCOVERS NEW FORM OF GRIEF • MISTRAL RAISES AT VALUATION OF GROSS DOMESTIC PRODUCT •
toolsWTF 5.7via Hacker News Frontpage

Which tools do Claude, Codex and Cursor choose? We measured 17k runs to find out

"Your AI agent's favorite tool is just 'ls -R' while it panics."

Explain Like I'm Normal

Armature analyzed 17,000 runs of top coding agents to see which CLI tools they actually invoke to solve tasks. The data reveals a massive preference for basic file navigation and grep over complex IDE features, highlighting a gap between how humans and agents navigate codebases. For founders, this provides a blueprint for which primitive tools are essential when building custom agentic workflows.

Read original ↗
#coding-agents#benchmarks#cursor#claude#software-engineering

GET THE DAILY CHAOS

The only newsletter for people who read AI news at 3am and feel things. One email a day.