Skip to content

#Agent

2 today

Jun 3Wednesday

Alibaba Technology · WeChat

Rethinking R&D Infrastructure When Agents Become First-Class Citizens

Xu Xiaobin argues that agent-based development compresses the intent-to-code loop from weeks or months to minutes, using a weekly-report system, a multi-role agent development setup, and image-repository provisioning as examples; the article identifies mismatches in Git, CI, code review, release flows, permissions, harness setup, and dry-run validation.

Why it matters: Xu Xiaobin argues from his own practice that AI agents cut the idea-to-working-code cycle from weeks or months down to minutes, with weekly-report systems and multi-role collaboration as concrete cases.

NVIDIA Blog

NVIDIA Partners With Microsoft on Unified Stack for Agentic AI Deployment

NVIDIA and Microsoft announced a unified agentic AI deployment stack at Build across Windows, Azure, and local environments; RTX Spark provides 1 petaflop of AI performance, while DGX Station for Windows offers 20 petaflops of FP4 performance and up to 748GB of coherent memory.

Why it matters: NVIDIA and Microsoft linked their AI deployment stack so one app runs on Windows devices, Azure cloud and local DGX workstations without code changes, alongside the 1 petaflop RTX Spark hardware.

Jun 2Tuesday

claude.dev Blog

How Claude Code dynamic workflows tailor an execution framework per task

Claude Code shipped dynamic workflows last week. Claude writes an execution framework for the current task on the spot, and workflows can be saved, shared and reused. They coordinate subagents through agent(), parallel() and pipeline() in JavaScript, can pick a model and a separate worktree, and resume a session after an interruption.

Why it matters: Claude Code splits execution and adversarial verification across subagents with separate contexts, a concrete orchestration method for cutting self-preference and goal drift on complex tasks.

May 28Thursday

NVIDIA Blog

NVIDIA Research Advances Robotics From Simulation to the Real World

NVIDIA Research presented 8 ICRA papers on sim-to-real robotics: ScheduleStream delivered a 3x speedup for multi-arm planning, COMPASS reached about 80% success across 20 real-world navigation trials, and Grasp-MPC achieved about 75% real-robot grasping success.

Why it matters: NVIDIA's COMPASS navigates real robots after simulation-only training, hitting roughly 80% success across 20 trials, 4.5x a pure imitation baseline.

Mistral AI

Mistral upgrades Le Chat into unified agent Vibe, covering office work and coding

Mistral upgraded Le Chat into a unified AI agent called Vibe, with one license covering both office work and coding. Existing chats, settings and plans all carry over. Work Mode supports enterprise knowledge search, structured data analysis, document and report generation, scheduled multi-step tasks and reusable skills, and connects to Google Workspace, Outlook, SharePoint, Slack, GitHub and more.

Why it matters: Mistral turns Le Chat into a unified agent called Vibe, with one license covering office work and coding while keeping existing chats, settings and plans, so current users are the ones to watch.

Alibaba Technology · WeChat

AI-Native Project Management: Two Git Repos Replace Weekly Updates, Insights, and Metrics Reports

Zhou Zhiwei describes a project-management setup that uses two Git repositories, an AI coding assistant, Shell, and Python to replace at least 80% of manual weekly-update chasing, data moving, chart generation, and engineering-metrics reporting.

Why it matters: Zhou Zhiwei used two Git repos plus AI coding assistants and scripts to automate weekly report chasing, data moving and performance reporting, offering a transferable way to rework project management.

Hugging Face Blog

ITBench-AA: Frontier Models Score Below 50% on the First Benchmark for Agentic Enterprise IT Tasks

Artificial Analysis and IBM published the ITBench-AA title, saying frontier models scored below 50% on an enterprise IT agent task benchmark; the post does not disclose tested models, sample size, or scoring method.

Why it matters: Artificial Analysis and IBM's ITBench-AA tests site reliability engineering tasks, and the top models, Claude Opus 4.7 at 47% and GPT-5.5 at 46%, both fall short of half.

May 27Wednesday

Alibaba Technology · WeChat

From Language Emergence to Collaborative Emergence: How AI Can Make High-Quality Decisions

Lv Ruofan proposes the Agent Room model: multiple agents share context, a task ledger, Memory, Runtime, and Artifacts, and two software-engineering cases show the system moving from workflow automation toward collaborative judgment rather than predefined task routing.

Why it matters: Alibaba technologist Lü Ruofan's Agent Room model has multiple AI agents share context and a task ledger, shown through two software engineering cases.

May 22Friday

Mistral AI

Mistral launches Connectors in Studio with built-in and custom MCP

Mistral launched Connectors in Studio. All built-in connectors and custom MCP are now callable through the API/SDK by every model and agent. New features include direct tool calling, human-in-the-loop approval flows, and programmatic access to create, modify, list and delete connectors.

Why it matters: Mistral registers built-in connectors and custom MCP servers in one place in Studio, with code samples for creating, calling and human approval, showing enterprise integration moving from repeated per-team work to one platform config.

May 21Thursday

Alibaba Technology · WeChat

Building an Agent from 0 to 1: Principles and Personal Assistant Practice

Zhan Xupeng published a roughly 50-minute article on Agent theory and a personal assistant implementation, covering memory, ReAct planning, progressive skill loading, subagents, and harness-level fault recovery.

Why it matters: Zhan Xupeng's long writeup breaks down agent memory, ReAct planning, skill loading and subagent division of labor, plus a path to building a personal assistant from scratch.

May 20Wednesday

claude.dev Blog

Using Claude Code: why HTML beats Markdown for complex output

Thariq Shihipar, a member of the Claude Code team, shares a method for generating complex output in HTML instead of Markdown, to improve reading, visualization, sharing and interaction. He exports interface tweaks as prompts so manual edits feed back into the agent workflow.

Why it matters: A Claude Code team member generates interactive documents in HTML and turns interface tweaks into prompts, feeding manual edits back into the agent workflow.

Alibaba Technology · WeChat

Zhenwu M890 AI Chip Debuts as Agentic Compute Foundation

Alibaba released a 128-card supernode server based on T-Head’s Zhenwu M890 AI chip, with P2P latency below 150 ns and rack bandwidth at the Pb/s level; it is live on Alibaba Cloud Bailian and supports Qwen, DeepSeek, and Kimi.

Why it matters: Alibaba's T-Head unveiled the Zhenwu M890 AI chip, with a 128-card supernode server already live on the Bailian platform supporting Qwen, DeepSeek and Kimi, showing where domestic AI chips land in agent workloads.

May 18Monday

Google DeepMind

Google DeepMind adds Street View grounding to Project Genie

Google DeepMind has added Street View real-scene grounding to its experimental prototype Project Genie. Users can pick a US location, then pair it with a style and characters to generate a world.

Why it matters: With Street View imagery wired in, agents and robots can train and navigate in virtual environments that track real places.

May 17Sunday

Google DeepMind

Google DeepMind launches Gemini for Science toolset

Google DeepMind released Gemini for Science, which includes three experimental tools on Google Labs: Hypothesis Generation, built on Co-Scientist.

Why it matters: DeepMind's research prototypes, including Co-Scientist and AlphaEvolve, become three tools open in Google Labs.

May 16Saturday

Google DeepMind

Google DeepMind releases Gemini 3.5 Flash

Google DeepMind released the Gemini 3.5 model family, with the first model, Gemini 3.5 Flash, available the same day in the Gemini app, Google Search AI Mode, Google Antigravity, the Gemini API and Gemini Enterprise.

Why it matters: Gemini 3.5 Flash beats Gemini 3.1 Pro on coding and agent benchmarks and outputs 4x faster than other frontier models, shipping the same day in the Gemini app, Search AI Mode and the Gemini API.

May 15Friday

Alibaba Technology · WeChat

Qoder 1.0 launches as an agentic development workspace beyond AI IDE

Alibaba released Qoder 1.0 with downloads for Windows, macOS, and Linux, adding a standalone Quest workspace, cross-project parallel agent tasks, a team knowledge engine, and Experts mode with five roles for planning, research, coding, review, and testing.

Why it matters: Alibaba's Qoder 1.0 adds a Quest workspace for running multiple agents in parallel across projects, plus a team knowledge engine that turns docs and code into searchable context.

May 13Wednesday

OpenAI News

Building a Safe, Effective Sandbox for Codex on Windows

OpenAI built a secure sandbox for Codex on Windows. The RSS snippet discloses controlled file access and network restrictions, but the post does not disclose implementation details, performance data, or rollout conditions.

Why it matters: An OpenAI engineer explains why Codex on Windows had no sandbox, leaving users to approve every command or grant full access, and how the team built one from scratch.

May 12Tuesday

Google DeepMind

Google DeepMind publishes Co-Scientist multi-agent research system

Google DeepMind published Co-Scientist research in Nature, introducing a Gemini-based multi-agent AI system that iteratively generates, debates and evolves new hypotheses for complex scientific problems.

Why it matters: Google DeepMind publishes Co-Scientist in Nature, a system of Gemini-driven specialized agents that generate, debate and evolve hypotheses, already applied to topics such as antimicrobial resistance.

May 8Friday

Alibaba Technology · WeChat

The AI-Native Era: Where R&D Organizations Go Next

Xu Xiaobin cites internal interviews showing that engineers who use AI heavily cut coding time from 30% to 5%, raised Agent conversation time from 5% to 60%, and increased end-to-end delivery efficiency by 2 to 3 times, while pure coding efficiency rose 10 times.

Why it matters: Alibaba's Xu Xiaobin says coding dropped from 30% to 5% of an engineer's time, with 60% now spent talking to agents.

NVIDIA Blog

Powering the Next American Century: Chris Wright and NVIDIA’s Ian Buck on Genesis Mission

The U.S. DOE and NVIDIA are building two AI supercomputers at Argonne; Equinox uses 10,000 Grace Blackwell GPUs. Solstice will use 100,000 Vera Rubin GPUs, which Buck said reach 5,000 exaflops. The key bottleneck is grid work: Wright said AI can cut interconnection studies from years to weeks or hours.

Why it matters: The US Department of Energy and NVIDIA are deploying two AI supercomputers at Argonne National Laboratory, scaling from 10,000 Grace Blackwell chips to 100,000 Vera Rubin GPUs.

May 6Wednesday

NVIDIA Blog

NVIDIA and ServiceNow Partner on Autonomous AI Agents for Enterprises

NVIDIA and ServiceNow expanded their partnership with Project Arc, an enterprise desktop agent. It connects via Action Fabric and uses OpenShell for sandboxed, policy-governed execution. Blackwell delivers over 50x Hopper’s token output per watt and nearly 35x lower cost per million tokens.

Why it matters: NVIDIA and ServiceNow built Project Arc, a desktop AI agent that reaches enterprise software through Action Fabric and runs permissioned actions inside an OpenShell sandbox.

May 5Tuesday

OpenAI News

GPT-5.5 Instant: smarter, clearer, and more personalized

OpenAI updated ChatGPT’s default model to GPT-5.5 Instant for default chat use. The RSS snippet says answers are more accurate, hallucinations are reduced, and personalization controls improved; the post does not disclose metrics, pricing, or context window.

Why it matters: OpenAI made GPT-5.5 Instant the default ChatGPT model, saying it cuts hallucinated content by 52.5% versus the previous generation on high-stakes medical, legal and financial questions.

May 1Friday

NVIDIA Blog

Nemotron Labs: What OpenClaw Agents Mean for Every Organization

NVIDIA says OpenClaw reached 250,000 GitHub stars by March 2026, passing React within 60 days. OpenClaw is Peter Steinberger’s self-hosted persistent agent; NVIDIA introduced NemoClaw with OpenShell sandboxing and Nemotron models. The key issue is governance: the post claims reasoning AI raised token use 100x, and autonomous agents add another 1,000x.

Why it matters: OpenClaw's self-hosted agents hit 250,000 GitHub stars in 60 days, faster than React, prompting NVIDIA to weigh token costs and safety limits for always-on agents.

Apr 30Thursday

Google DeepMind

Google DeepMind announces AI co-clinician medical research program

Google DeepMind announced an AI co-clinician research program, exploring how AI agents can assist patient care under a doctor's clinical supervision. In a blinded evaluation of 98 real primary care queries, the system made no critical errors in 97 cases, and doctors preferred its answers over mainstream evidence synthesis tools. On 140 consultation skills, it matched or beat primary care physicians on 68, but expert physicians were still better overall at spotting red flags and key physical exams.

Why it matters: Google DeepMind details its AI co-clinician research program: in a blinded review of 98 real primary care queries, 97 had no critical errors, and doctors preferred its answers over mainstream evidence synthesis tools.

claude.dev Blog

Building Claude Code: designing agents around prompt caching

The Claude Code team shared engineering lessons on prompt caching, keeping the request prefix stable to cut latency and cost in long agent sessions. Put static content first, update state through messages, avoid switching models or adding and removing tools mid-session, and use defer_loading to load tools late. When compacting context, reuse the original session's system prompt, tool definitions and history, then append the compaction request so it hits the cache. The team also tracks cache hit rate and treats an abnormal drop as an incident.

Why it matters: The Claude Code team applies cache prefix stability to prompt ordering, tool loading and context compaction, a transferable caching method for long agent sessions.

Apr 29Wednesday

NVIDIA Blog

NVIDIA Launches Nemotron 3 Nano Omni for Vision, Audio, and Language Agents

NVIDIA launched Nemotron 3 Nano Omni, claiming up to 9x higher throughput at the same interactivity. It uses a 30B-A3B hybrid MoE with Conv3D, EVS, and 256K context, taking text, images, audio, video, documents, charts, and GUIs as input. Open weights, datasets, and training methods arrive April 28, 2026 on Hugging Face, OpenRouter, build.nvidia.com, and 25+ platforms.

Why it matters: NVIDIA open-sourced Nemotron 3 Nano Omni, one model handling image, text, audio and video with 3B active parameters and 256K context, giving teams building multimodal agents an open path.

Apr 28Tuesday

X · @claudeai

Claude Now Connects to Tools Creative Professionals Already Use

Claude added a Blender connector for scene debugging, tool building, and batch object edits from Claude. The post does not disclose versions, pricing, or rollout scope; the key issue is agent control boundaries inside DCC workflows.

Why it matters: Claude now has a Blender connector that can debug scenes, write tools or batch-edit objects in conversation, showing what an AI assistant can do inside professional 3D software.

Apr 27Monday

Mistral AI

Mistral AI opens public preview of Workflows

Mistral AI has put Workflows, its enterprise AI orchestration layer, into public preview. It offers durable execution, observability and human-in-the-loop approvals. ASML, ABANCA and CMA-CGM are already using it to automate critical processes.

Why it matters: Mistral AI opens Workflows, its enterprise AI orchestration layer, in public preview, with ASML and ABANCA already automating key processes and material showing durable execution, observability and human approval in freight release and KYC review.

OpenAI News

An Open-Source Spec for Orchestration: Symphony

OpenAI released Symphony, an open-source spec for Codex orchestration. The RSS snippet says it turns issue trackers into always-on agent systems; the post does not disclose spec details, license, APIs, or benchmarks.

Why it matters: OpenAI open-sourced the Symphony orchestration spec, turning a Linear board into a dispatch point for coding agents where humans only review results; internal teams merged five times as many PRs in three weeks.

Apr 25Saturday

X · @AnthropicAI

New Anthropic research: Project Deal

Anthropic announced Project Deal and had Claude buy, sell, and negotiate for employees in a San Francisco office marketplace. The setup is confirmed as an internal marketplace; the post does not disclose scale, model version, or outcome metrics.

Why it matters: Anthropic built an internal trading market in its San Francisco office where Claude buys, sells and haggles for colleagues, a look at how models handle real person-to-person transactions.

Apr 24Friday

Hugging Face Blog

DeepSeek-V4: a million-token context that agents can actually use

DeepSeek released V4 with two MoE checkpoints, Pro and Flash, both supporting a 1M-token context. Pro has 1.6T total and 49B active parameters; Flash has 284B total and 13B active. The key detail is KV cost: Pro uses 27% of V3.2 single-token FLOPs and 10% of its KV cache; Flash uses 10% and 7%.

Why it matters: DeepSeek-V4 comes in Pro and Flash, both with a million-token context, Pro at 1.6 trillion total parameters and 49 billion active, Flash at 284 billion total and 13 billion active.

X · @claudeai

Memory on Claude Managed Agents is now in public beta

Claude has put Memory for Managed Agents into public beta, and agents can now learn from every session. The post only says it uses an intelligence-optimized memory layer balancing performance and flexibility; it does not disclose capacity, retention, pricing, or access conditions. What matters for practitioners is when persistent memory becomes default and how it changes agent evals and state management.

Why it matters: Claude managed agents gained a memory layer that keeps context across conversations, with Anthropic saying it traded off performance against flexibility, which affects how agents are evaluated and how state is managed once they persist.

X · @claudeai

Claude can now connect to more apps outside work, including Tripadvisor, Booking.com, and Resy

Claude added at least 10 consumer app connections, including Tripadvisor, Booking.com, Resy, Instacart, Spotify, Audible, AllTrails, Thumbtack, and TurboTax. The RSS snippet confirms only a product update; the post does not disclose integration method, supported actions, regions, permission scope, or rollout timing. The key question is whether Claude can act in these apps directly, not just list them.

Why it matters: Anthropic added about a dozen consumer app connections to Claude, including TripAdvisor, Booking and Resy, pushing the assistant past office work into dining, travel and shopping.

X · @OpenAI

Introducing GPT-5.5

OpenAI introduced GPT-5.5, and it is now available in ChatGPT and Codex. The RSS snippet says it targets real work and agents, can understand complex goals, use tools, check its work, and carry more tasks to completion; the post does not disclose parameters, pricing, context window, or benchmark results. What matters is the execution loop, not the headline's “new class of intelligence.”

Why it matters: OpenAI put GPT-5.5 into ChatGPT and Codex, pitched at real work and agent tasks that run from start to finish.

Apr 23Thursday

Hugging Face Blog

How to Use Transformers.js in a Chrome Extension

Hugging Face published a guide for a Transformers.js Chrome extension using Gemma 4 E2B. It defines three MV3 entry points: background service worker, side panel, and content script. The key design keeps local inference in the background and uses messaging plus a tool loop.

Why it matters: Hugging Face walks through the engineering trade-offs of running Gemma 4 E2B locally in a Chrome MV3 service worker.

X · @OpenAI

Introducing workspace agents in ChatGPT—shared agents for complex tasks and long-running workflows

OpenAI announced workspace agents in ChatGPT, described as shared agents that work across tools and teams for complex and long-running workflows. Only the title and RSS snippet are disclosed; the post does not disclose supported tools, pricing, access tier, permission model, or rollout timing. The key issue to watch is the collaboration boundary of shared agents, not the headline claim alone.

Why it matters: OpenAI launches workspace agents in ChatGPT as shared team assistants for complex tasks and long-running workflows, and the collaboration boundaries and permission design are key to how shared agents get deployed.

Apr 22Wednesday

OpenAI News

Introducing workspace agents in ChatGPT

OpenAI introduced workspace agents in ChatGPT, describing them as Codex-powered agents that automate complex workflows in the cloud. The RSS snippet confirms secure work across tools for teams, but the post does not disclose pricing, availability, supported tools, or performance metrics.

Why it matters: OpenAI is putting Codex-powered workspace agents into paid ChatGPT plans, so team automation moves from chat to running multi-step tasks across tools in the cloud.

OpenAI News

Speeding up agentic workflows with WebSockets in the Responses API

OpenAI says WebSockets in the Responses API speed up the Codex agent loop, using connection-scoped caching to cut API overhead and improve latency. The RSS snippet confirms the mechanism, but the post does not disclose latency deltas, throughput numbers, or workload conditions. The key point is transport-layer optimization, not a new model.

Why it matters: OpenAI added WebSocket long connections and connection-level caching to the Responses API, targeting the case where faster inference makes API overhead the bottleneck in agent loops, cutting repeated connection setup and duplicate requests.

Apr 17Friday

Tencent Technology · WeChat

From Vibe Coding to Agentic Engineering: Rebuilding the Full Backend Development Workflow

Tencent engineers report a one-week practice that used Claude Code plus custom Skills, Commands, and MCP servers to run an 11-stage backend workflow in one terminal session. The post gives reproducible details: one requirement-exploration step used 20 tool calls, 93.8k tokens, and 56 seconds; execution was split into 4 tasks and produced 3 commits. The real point is workflow orchestration, not raw code generation; human review remains at plan, deploy, and review gates.

Why it matters: A Tencent engineer wired 11 backend development steps into one terminal pipeline using Claude Code with custom Skills, Commands and MCP servers, and reports tool-call counts and token spend for the requirements phase.

X · @OpenAI

Codex for (almost) everything.

OpenAI said Codex can now use apps on Mac, connect to more tools, and handle ongoing and repeatable tasks. The post also claims image creation, learning from prior actions, and remembering user preferences; it does not disclose app coverage, integration method, pricing, or rollout timing.

Why it matters: OpenAI says Codex can now control Mac apps, plug into more tools and take over repetitive tasks, widening its role from writing code to operating the desktop.