AI Briefing#ai-agents
The latest AI and developer news about #ai-agents, with the original source and a short summary.
The latest AI and developer news about #ai-agents, with the original source and a short summary.
To improve an agent's accuracy and reliability, you need to build targeted evaluation systems that reflect real production behavior.
Agent middleware allows you to customize the core loop of an agent harness and inject business logic.
AI innovation is achieved through an ecosystem in which open source and proprietary models coexist and complement each other.
LangSmith Fleet has launched a 'Skills' feature that gives agents domain knowledge and allows it to be shared within teams.
OpenAI has unveiled a 'Safety Bug Bounty' program to identify AI misuse and safety risks.
Moda used Deep Agents and LangSmith to build a production-grade multi-agent system that performs complex visual design.
EVA, a new framework that simultaneously evaluates the accuracy and conversational experience of voice agents, has been released.
LangChain is participating in Google Cloud Next 2026, unveiling the latest technology for agent development and LangSmith updates.
LangSmith Fleet has introduced two agent authorization models: the on-behalf-of-user method and the dedicated-credentials method.
NVIDIA has unveiled OpenShell, a sandbox-based runtime, to strengthen the security of autonomous AI agents.
Agent 4 turns ideas directly into products through parallel work, collaboration, and Instant branching.
LangChain has launched LangSmith Fleet, an enterprise workspace that supports agent creation through management.
Pinterest runs agent automation with internal MCP servers and a registry.
Chat SDK deploys agents across multiple chat platforms like Slack, Teams, and Discord from a single codebase.
It's a dedicated infrastructure service that supports AI agents in independently sending and receiving emails.
Durable delivers AI agent-based business automation services to 3 million customers with just 6 engineers.
Meta's REA autonomizes ad ranking ML experiments, boosting accuracy and productivity.
At GTC, NVIDIA expands the AI agent ecosystem by unveiling RTX PCs and DGX Spark that run the latest open models locally.
OpenAI's **GPT-5.4 Mini** and **Nano** models have been added to **Vercel AI Gateway**, supporting agentic workflows.
Vercel has updated its Terms of Service to support AI agentic features and improve the platform.
This presents a way to centrally and efficiently manage scattered AI assets such as prompts and skills, and keep them up to date, through **MCP stdio**.
Forge trains, aligns, and evaluates frontier-level AI models using a company's internal knowledge.
A feature has been launched that lets you package Manus workflows into reusable agents called 'Skills' and share them with your team.
Vercel has selected 39 early-stage AI startups from around the world to run a 6-week accelerator program.