Improving agents The old way: Manually reading traces, looking for patterns, writing evals, and creating fixes. The better way: Letting LangSmith Engine run that cycle for you
AI
-
Qwen-VLA: Unified Vision-Language-Action Robot Learning
By
–
“Qwen-VLA: Unifying VLA Modeling across Tasks, Environments, and Robot Embodiments” They turned robot learning into one vision-language-action modeling problem instead of separate policies for each task, environment, and robot body. So by adding a DiT flow-matching action
-
Gemini Embedding 2: Native Multimodal Embedding Model
By
–
"Gemini Embedding 2" This paper turns Gemini into one native embedding model for text, image, video, audio, and interleaved multimodal inputs. Instead of converting everything into text first, it embeds raw modalities directly into one shared space, improving audio search,
-
Claude Code Dynamic Workflows and Opus 4.8 Explained
By
–
Claude Code Dynamic Workflows, explained! Anthropic dropped Opus 4.8, and everyone is talking about the benchmarks, the honesty improvements, and the cheaper fast mode. But the feature that shipped alongside it might matter more for how we actually build: Dynamic Workflows in
-
Agents get their own execution layer with ‘ego lite’
By
–
Chrome puppets are about to get buried.
— God of Prompt (@godofprompt) 29 mai 2026
ego lite is what happens when you stop forcing agents into browsers built for humans and give them their own execution layer.
Real login state. Background isolation. Complete browser control.
The toy-agent era is ending. https://t.co/TkayhTFQ4bChrome puppets are about to get buried. ego lite is what happens when you stop forcing agents into browsers built for humans and give them their own execution layer. Real login state. Background isolation. Complete browser control. The toy-agent era is ending.
-
Claude Code Acts as an AI Agent to Find Jobs on LinkedIn
By
–
CLAUDE CODE CAN NOW FIND YOU A JOB (literally)
— Nico (@nicos_ai) 29 mai 2026
Built with ego lite, the browser your agent drives directly.
I gave Claude Code one prompt:
→ Search LinkedIn for AI engineering roles in San Francisco
→ Full-time, on-site, posted in the last month
→ Exclude Senior, Staff,… https://t.co/kZdNlLfNMR pic.twitter.com/4TeGRBBGpfCLAUDE CODE CAN NOW FIND YOU A JOB (literally) Built with ego lite, the browser your agent drives directly. I gave Claude Code one prompt: → Search LinkedIn for AI engineering roles in San Francisco
→ Full-time, on-site, posted in the last month
→ Exclude Senior, Staff, -
Agent Safety: Classifier Subagent for Tool Call Approval
By
–
Agent actions that aren't on your allowlist or can't be sandboxed go to a classifier subagent. This separate agent decides whether to allow the tool call, try a different approach, or ask you for approval. Learn more:
-
Cursor Auto-Review Mode Enables Safer Agent Tool Execution
By
–
Auto-review mode is now available in Cursor.
— Cursor (@cursor_ai) 29 mai 2026
It allows agents to run tool calls with fewer approval prompts and safer execution. pic.twitter.com/GZQX89mgmqAuto-review mode is now available in Cursor. It allows agents to run tool calls with fewer approval prompts and safer execution.
-
Open AI Models Gaining Traction Among AI Teams
By
–
The latest finding in the LangSmith Signal: Open Models are having a moment. 1 in 3 AI teams ran an open-weights model in April 2026, up from 1 in 5 nine months ago. The overall number of teams using open weights grew 3x. We’re seeing newer users choose open models at a
-
Grok AI Model Praised for Agentic Capabilities and Interface Design
By
–
grok-build-0.1 for interface design.
— Pietro Schirano (@skirano) 29 mai 2026
Pretty impressed with the agentic capabilities and multi tool calling.
good model. pic.twitter.com/XGvX2C7o9Jgrok-build-0.1 for interface design. Pretty impressed with the agentic capabilities and multi tool calling. good model.