Every article drawn from current research and sources.
Prime Intellect has released Prime Agent, an open-source coding harness that treats its own prompts, memory, and sub-agents as editable state — and claims a benchmark score that matches reported human expert performance on ARC-AGI 3.
Starting August 14, Anthropic is flipping the default for new Claude Code sessions on Pro, Max, and Team plans from manual approval to auto mode. The change is backed by a controlled study of 1,053 paid professional testers and a 720-attempt prompt-injection evaluation against Claude Code v2.1.205.
Gemini 3.7 Flash arrived on August 13, 2026, just three weeks after Gemini 3.6 Flash, bringing measurable gains in coding and agentic workflows across every benchmark Google published.
