WHAT NEW IN X10.29 THU

AI collections from X · 22 collections · 58,925 posts

Back to No.001 GPT-6 Astra in Practice: Tests, Use Cases, and FeedbackUpdated 13 hours ago
gpt-6IMAGE

Cline + GPT-6 Astra scores 23/29 on Composio's stateful tool-use benchmark

Composio ran Cline through its multi-app stateful tool-use benchmark, and Cline with GPT-6 Astra passed 23 of 29 tasks, the best Astra result among harnesses tested.

Composio
@composioArchived 10 days agoOriginally posted 2026.10.16
Cline + GPT-6 Astra scores 23/29 on Composio's stateful tool-use benchmark
Composio @composioOriginally posted 2026.10.16
Open original on X

We ran Cline through Composio Benchmark, an eval that measures stateful tool use across workflows with multiple connected apps. Cline + GPT 6 Astra passed 23/29 - the highest score of any harness we’ve tested with Astra.

x.com/composio/status/2111091002492510710Archived 2026.10.18 16:00
Open original on X