
Agents9 min read
Claude performed best on a new benchmark for ‘agents that build agents’. But it passed fewer than a quarter of the tests.
The New Stack (AI)
Every AI story we track on Sierra — 5 stories so far, each summarized in our own words and linked back to the publisher that reported it.




