GrubMarket launches AI agent for food distributor purchase orders
GrubMarket released an AI agent that transforms purchase order processing for food distributors, per AndNowUKnow.
GrubMarket released an AI agent that transforms purchase order processing for food distributors, per AndNowUKnow.
Why it matters: Export-control infighting directly shapes how much Nvidia AI compute reaches China.
Axios reports Commerce Secretary Howard Lutnick again stalled Trump-authorized curbs on companies suspected of diverting Nvidia AI chips to China, with Bessent confronting him over not labeling a customer a security threat.
Why it matters: Shared effort to harden agent security as prompt injection remains a top AI risk.
Red Hat and Nvidia developed OpenShell to defend AI agents against prompt injection attacks, according to SiliconANGLE.
SambaNova says Z.ai's open-weight GLM-5.3 model is available on SambaCloud inference at more than 200 tokens per second, reachable via playground, API and coding-agent integrations.
Why it matters: A major travel merchant drawing a line against agentic booking tests the limits of agent commerce.
United Airlines moved to block third-party AI agents from booking flights and accessing customer accounts ahead of the holidays, citing privacy and security concerns.
A developer released a local world model that turns any image into an interactive video you control in real time, running on RTX 3060 and 4060 GPUs or above, hosted as lucidml/gta on Hugging Face.
J.Crew launched an AI shopping tool designed to think like a stylist and recommend outfits to shoppers.
Sorce, a Y Combinator Fall 2025 startup, lets users swipe right on jobs and sends an AI agent to submit applications. It says 850,000 people have used the tool since its August 2024 launch.
eBay expanded its seller 'magical listing tool' with a workflow that turns a user's own photo directly into an AI-generated listing, Digital Commerce 360 reports.
Why it matters: A major cloud vendor shipping codebase written entirely by autonomous agents tests how far agentic coding has come.
AWS Emulator is a downloadable local emulator for AWS services. AWS says thousands of autonomous AI agents wrote every line of its code, offering a test case for agent-written software.
Why it matters: A calibrated decision-focused API signals Google pushing Gemini beyond chat into structured agent workflows.
Google released the Gemini Decisions API, powered by its Gemini Decision model, for fast, calibrated multimodal decisions and structured outputs.
OpenEvidence released Access Intelligence, a suite built into its clinical AI platform that helps clinicians with affordability and coverage tasks like eligibility and availability information. It is free to all clinicians, available immediately.
Why it matters: Giving agents callable APIs instead of browser clicking could make agent tool use faster and more reliable.
Pomerado, a YC F26 startup, converts any website into an MCP endpoint so AI agents can call an API instead of clicking through pages.
Decagon introduced Alto, a product for controlling how voice AI agents conduct conversations, Unite.AI reports.
Why it matters: Shows agents can automate model R&D cycles, compressing weeks of experimentation into hours.
Salesforce introduced SFR-AutoR&D on November 11, 2026. Its agents produced the SFR-Embedding v4 model in about 80 hours instead of the several weeks human experimentation often requires, by autonomously changing codebases and running the research cycle.
Unsloth published Dynamic v3.1 GGUF builds of Qwen3.8-27B, claiming 5-10% better accuracy on DeepSWE and Terminal Bench than earlier quantizations; the local-run format targets consumer hardware.
Legal software provider LawVu said five legal AI tools — Anthropic's Claude, Microsoft Copilot, OpenAI's ChatGPT, Harvey and Legora — can now connect to its LegalOS platform with no custom development.
Klad is a source-available paint app designed so a coding agent can operate it, letting agents create and manipulate images programmatically.
Tabletop is a Google Labs experiment where groups build interactive itineraries, polls and brackets together. Users can join a waitlist.
DNSFilter made its biggest platform expansion yet, adding AI and threat-intelligence features, according to its CEO, aimed at AI-powered threats and unified security controls.
Nvidia received a downgrade amid uncertainty around Jensen Huang, who pulled out of Anthropic and ElevenLabs events after an affair admission, as the company investigates.
Why it matters: Military targeting AI raises questions about automation speed versus human control.
Helsing's Clarity uses agents to distill sensor data into target recommendations for weapons systems, with every decision attributed to a human and under European control.
Firefox 160 adds on-device speech-to-text and support for MKV files with FLAC, MP3 and PCM audio, expanding browser-native media and voice capabilities.
London- and California-based Claren launched Envoys, always-on agents that pick up in-house legal tasks from email or Slack and run the matter end to end.
Why it matters: Business context and ontology are pitched as the key to reliable autonomous enterprise agents.
Databricks launched Genie Ontology and a revamped Genie, giving AI agents business context across more data sources. The company claims 1.6x accuracy using less than half the tokens.
Why it matters: Cell-type-specific variant prediction could sharpen how genomics links DNA changes to disease.
AlphaGenome 2 is a DeepMind AI model that predicts how DNA sequence changes affect specific human cell types, aiming to explain disease at the cellular level.
Atuin Threads is a tool that collects AI agent sessions from every machine so you can view and resume them in one place.
Bengaluru Police launched Voice AI-112, a voice AI emergency response system that has gathered over 150,000 responses.
Why it matters: Standardizes how agents render UI across platforms, potentially a key agentic interface layer.
A2UI 1.0 is an open protocol for generative UI that lets AI agents render interactive native screens across web, mobile and desktop, with open-source renderer SDKs for React, Angular, Lit, Flutter, SwiftUI and Jetpack Compose and agent SDKs.
VIU by HUB, a digital insurance brokerage backed by HUB International, launched a ChatGPT plugin letting consumers in 48 states compare home and auto insurance quotes, with a licensed advisor following up to bind coverage.
Why it matters: Adobe's experiments signal generative video editing moving toward real-time, conversational workflows.
At Adobe Max, Adobe showed experimental tools: turning photos into explorable 3D scenes, designing AI characters by talking to them, and a real-time generative AI video editing feature.
Edelbyte is an Israeli data-infrastructure company building a search and analytics platform for AI-driven demand at much lower cost than legacy systems. It emerged from stealth with $70 million.
Valory unveiled Olas-Predict-Qwen3.8-27B, an open-weight 27B forecasting model it says comes within 1.2 points of GPT-5.6 on real-world prediction markets, reaching 67.3% balanced accuracy in internal tests.
Why it matters: A named delivery brand building directly into a chatbot shows agentic commerce moving from pilots to consumer interfaces.
Gopuff launched a fully integrated delivery experience inside xAI's Grok Bot, letting users order through the chatbot; Grok Bot 1.0 adds Slack integration and video call support.
SAP and VF Corporation opened S.Fashion, an AI-enabled retail innovation space at SAP headquarters in Walldorf, stocking The North Face, Vans and Timberland to test retail AI processes end to end.
Sangoma launched Voice AI Agents aimed at winning callers, adding AI voice handling to its communications products.
Nameshot is a Mac app that automatically renames screenshots on-device using Apple Intelligence, so files are searchable instead of carrying timestamps. Built by an Apple Design Award-winning game designer.
Emirates NBD launches real-time AI market insights for digital wealth customers, adding Bridgewise AI market insights to its ENBD X app.
Why it matters: Gemini reaching CarPlay puts Google's assistant in front of iPhone drivers, a notable cross-platform move.
Google is bringing its Gemini AI app to Apple CarPlay for iPhone users, expanding the assistant beyond Android, with related Gemini updates on Pixel and Motorola phones.
Why it matters: Longer AI clips and outside access widen AI video creation inside a major social platform.
TikTok extended AI video generation in Symphony to 30 seconds and opened paid BytePlus access to the tool for outside developers and marketers.
Zipline says its delivery drones now use onboard sound modeling to optimize flight paths for quietness in real time, an industry first for autonomous drone delivery. Full rollout details weren't given.
Why it matters: Explosion-proof quadrupeds expand inspection robots into hazardous energy and chemical plants.
Swiss robotics firm ANYbotics has begun delivering ANYmal X, a four-legged inspection robot built for potentially explosive environments, to industrial customers including Equinor and DSM-Firmenich for oil, gas and chemical sites.
STMicroelectronics' STM32G5 series pairs a 280 MHz Arm Cortex-M33 core with mixed-signal peripherals, up to 512KB flash and 112KB SRAM, targeting digital power, motor control, robotics and EV chargers.
Horkos and Spain's Meltio 3D launched the NX70 Duo version AM, a twin-head machining center combining wire laser metal additive manufacturing with high-precision cutting in one machine.
MSI's MEG X is a 34-inch ultrawide gaming monitor with Samsung's Penta Tandem QD-OLED panel at 360Hz, on-device AI features and next-gen connectivity. Price and availability not yet announced.
Blue Phone is a $1,299 Pixel 10 Pro offering an AI assistant pitched as a chief of staff that goes beyond Gemini, aimed at switching iPhone users.
Alpamayo 2 Edge is Nvidia's open self-driving model that takes in six cameras and recent motion, trained in an RL loop. Embedl released an optimized build over 10x faster on AGX Thor.
Why it matters: Infrastructure leadership churn at a top lab during an unprecedented AI capex cycle.
Meta's longtime head of infrastructure, Santosh Janardhan, will leave the company in early 2027, a major leadership change amid its hundreds-of-billions AI and data center buildout.
Matrix Robotics' MATRIX-3 humanoid appeared as a tour guide at China's CIIE trade show, its first real-world deployment in that role.
AWS Marketplace now supports usage-based metering for Amazon SageMaker AI bidirectional streaming inference, letting model providers charge for what customers consume. Deepgram is the launch partner. AWS also detailed UAEF, a framework-agnostic system for closed-loop AI agent evaluation.
Israeli defense firm Rafael launched a new cruise missile that uses AI assistance, part of a growing wave of AI-enabled military systems.
Google's first big software update for its Googlebook laptops frees 3GB to 6GB of system RAM that on-device AI features were holding, per TechSpot.
Meta is rolling out a software fix for its AI glasses after warning that some models may become too warm when exposed to moisture.
Hitachi Vantara's VSP One Object can now query supported open-format data directly on the storage platform, with a four-node starting configuration; it also adds governance features to Hitachi iQ Studio for enterprise AI.
Supermicro is now shipping next-generation H15 servers powered by 6th Gen AMD EPYC 9006 series CPUs, aimed at cloud, HPC and agentic AI workloads.
OmniOps, a Saudi sovereign AI infrastructure company, launched Bunyan Inference Hub, giving governments, enterprises, startups and developers access to open-source AI models and compute hosted in Saudi Arabia.
CoreWeave Database is generally available, offering managed PostgreSQL on CPU nodes inside customers' CoreWeave Kubernetes clusters alongside AI training and inference workloads.
Why it matters: A major appliance maker joins the race to put agentic humanoid-style robots in homes.
LG introduced CLOiD, an AI home robot it says can autonomously manage household tasks, unveiled at CES 2026.
Memories.ai offers standardized egocentric video collection plus AI-powered annotation, building training-data infrastructure for robotics developers.
Why it matters: A large round for an AI chip challenger shows investors still backing Nvidia alternatives.
London-based Fractile, which designs custom semiconductors for AI, raised $720M in a Series C with UK state backing at a $7 billion valuation.
Why it matters: Opens a real robot delivery fleet to outside developers and AI agents.
Serve Robotics launched an on-demand delivery API letting third parties and AI agents build new delivery services on its fleet of more than 2,000 sidewalk robots. Live access requires approval.
Microsoft's AI upscaling for games is now officially available on select Xbox Ally handhelds and some Copilot+ PCs, improving frame rates and image quality.
Why it matters: Coding agents that program industrial robots without training data could cut robotics deployment costs.
UC Berkeley and Nvidia released open-source tools for teams of AI coding agents that write, simulate and refine robot programs without training data.
SwitchBot, which makes AI-enabled home robotics, launched Curtain Track 4, an automated motor for existing U-rail curtains, priced from $109.99 with a Battery+ version rated for up to 12 months per charge.
Etisalat UAE now sells Meta's AI glasses starting at AED 1,160, bringing the wearable Ray-Ban-style glasses to the Emirati market.
Zed is a multiplayer code editor built for speed with AI-assisted coding features. Version 1.28.1 is a patch release for macOS, Windows and Linux.
Red Hat Enterprise Linux 10.3 and 9.9 make AI-assisted diagnostics a supported production feature, adding OS integrity controls, AI monitoring dashboards, developer updates and Model Context Protocol support.
AWS describes an event-driven pipeline using CloudTrail, EventBridge, Lambda and AgentCore Identity that detects MCP server changes and pushes updated tool metadata to the AWS Agent Registry without stored secrets, across single- and multi-account setups.
Amp, a CLI-based frontier coding agent and development environment, released build 0.0.1794427236-ga7dfca, installable via npm as @ampcode/cli.
OpenAI's personal finance features for ChatGPT are rolling out to the Mac desktop app, following their earlier launch elsewhere.
In iOS 27, higher iCloud+ storage tiers gain AI features for HomeKit cameras as a new perk, 9to5Mac reports.
AWS's Generative AI Innovation Center published a catalog of 23 production-ready, open-source reusable assets covering RAG pipelines, document processors and agent scaffolding, with delivery results.
xAI's Grok Bot is now available in Microsoft Teams, letting users bring the Grok assistant into team chats. Details on setup and features are documented on docs.x.ai.
Yogurt is a free, open-source TypeScript app that takes private AI meeting notes, keeping transcripts and summaries local rather than sending them to a hosted service. It drew 133 GitHub stars in its first week.
A Google Home update brings back classic Assistant sounds for special timers, changes the smart-home confirm sound, and fixes Nest Live View issues while improving Nest Hub and Gemini for speakers.
LASIK is a static CLI and MCP server that reverses Blind LLM Agents hardening so coding agents can read real code again, without executing the target. It reached 100 GitHub stars in its first week.
Sorce, a Y Combinator Fall 2025 startup, lets users swipe right on jobs and sends an AI agent to submit applications. It says 850,000 people have used the tool since its August 2024 launch.
LLM, a CLI for running prompts against language models, ships 0.37.2 fixing two security issues in extra-openai-models.yaml handling, now managed by the llm openai endpoint add command.
Antigravity, Google's agent orchestration platform for coding, releases version 2.31.1 for macOS, Windows and Linux, adding a /swarm command that lets coding agents debate on a shared board.
Why it matters: A major cloud vendor shipping codebase written entirely by autonomous agents tests how far agentic coding has come.
AWS Emulator is a downloadable local emulator for AWS services. AWS says thousands of autonomous AI agents wrote every line of its code, offering a test case for agent-written software.
Why it matters: A calibrated decision-focused API signals Google pushing Gemini beyond chat into structured agent workflows.
Google released the Gemini Decisions API, powered by its Gemini Decision model, for fast, calibrated multimodal decisions and structured outputs.
Why it matters: Giving agents callable APIs instead of browser clicking could make agent tool use faster and more reliable.
Pomerado, a YC F26 startup, converts any website into an MCP endpoint so AI agents can call an API instead of clicking through pages.
Sophon-Realtime v1 is an open, Apache-2.0 full-duplex voice model that listens continuously, picks up when you stop talking and lets you interrupt, while sensing tone such as hesitation or sarcasm.
Why it matters: Shows agents can automate model R&D cycles, compressing weeks of experimentation into hours.
Salesforce introduced SFR-AutoR&D on November 11, 2026. Its agents produced the SFR-Embedding v4 model in about 80 hours instead of the several weeks human experimentation often requires, by autonomously changing codebases and running the research cycle.
Legal software provider LawVu said five legal AI tools — Anthropic's Claude, Microsoft Copilot, OpenAI's ChatGPT, Harvey and Legora — can now connect to its LegalOS platform with no custom development.
AWS explains how to build an AI-driven product development lifecycle with six AWS Quick agents that move from scattered customer signals to a validated prototype across Discover, Decide, Design and Prototype, with human approval at each gate.
Klad is a source-available paint app designed so a coding agent can operate it, letting agents create and manipulate images programmatically.
Tabletop is a Google Labs experiment where groups build interactive itineraries, polls and brackets together. Users can join a waitlist.
As AI agents become the main enterprise workload, Google's agent sandboxing aims to give each agent hard walls without runaway compute costs while they sit idle.
Firefox 160 adds on-device speech-to-text and support for MKV files with FLAC, MP3 and PCM audio, expanding browser-native media and voice capabilities.
Why it matters: Business context and ontology are pitched as the key to reliable autonomous enterprise agents.
Databricks launched Genie Ontology and a revamped Genie, giving AI agents business context across more data sources. The company claims 1.6x accuracy using less than half the tokens.
A Gradio demo for text-to-image and image editing based on MiniMax-H3, with prompt upsampling, aspect-ratio and resolution controls, up to 8 reference images and before/after view.
Atuin Threads is a tool that collects AI agent sessions from every machine so you can view and resume them in one place.
Why it matters: Standardizes how agents render UI across platforms, potentially a key agentic interface layer.
A2UI 1.0 is an open protocol for generative UI that lets AI agents render interactive native screens across web, mobile and desktop, with open-source renderer SDKs for React, Angular, Lit, Flutter, SwiftUI and Jetpack Compose and agent SDKs.
Plurai raised $13M in seed funding and opened a platform that simulates and crash-tests AI agents before they reach customers, using self-driving-style end-to-end training.
Sophon-Realtime v1 is an open, Apache-2.0 full-duplex voice model that listens continuously, picks up when you stop talking and lets you interrupt, while sensing tone such as hesitation or sarcasm.
A Gradio demo for text-to-image and image editing based on MiniMax-H3, with prompt upsampling, aspect-ratio and resolution controls, up to 8 reference images and before/after view.
A developer released code and a paper for training a 27B model on a single 24GB RTX 3090, requiring hundreds of billions of tokens for good quality, alongside converted model repos and a 40.4GB Qwen GGUF.
A Hugging Face Space runs an uncensored version of the Qwen-Image-2.1 image model for generating and editing images, using a LoRA on the official Diffusers weights with ZeroGPU.
ZONOS3-Mini is an open-source text-to-speech model with 100M active parameters, released by Zyphra on Hugging Face for generating spoken audio from text.
MiniMax released H3-Image, an open-weight 16-megapixel text-to-image and editing model on Hugging Face that runs in six steps with 4K output.
Propaganda_Eval is a Hugging Face dataset of 803 prompts on topics China censors, each with required facts; answers are scored on factuality and adoption of state narratives.
Mirage is an Apache-2.0 image-text-to-text model that describes scenes, detects objects, reads text, segments images and extracts structured data. Weights are on Hugging Face with a Python package.
Arete is a Hugging Face dataset of prompt-response pairs on seven values topics, each answered by an Inkling teacher model and the Qwen 3.8-27B checkpoint, with category tags and verification hashes.
GGUF quantized versions of the Gemma-4-31B-it-ARA-heretic model, for running locally with tools like llama.cpp. Static and imatrix quants are available.
A new paper introduces MMDM, a benchmark measuring massive multitask decision making for System-1 decision models that make structured probabilistic decisions in a single forward pass, to test generalization.
A new paper introduces DexForesight, a framework that learns spatial foresight implicitly by injecting dynamics-aware world representations into the latent space of a dexterous manipulation policy instead of predicting futures explicitly.
A research paper finds LLM agent memory can cause degradation when agents misapply lessons drawn from past trajectories, and proposes evidence-linked memory that ties each lesson to its source to reduce failures.
The paper casts weight-space repair of language models as sparse discrete optimization, editing bits to fix reasoning loops without gradients, extending prior agentic weight-programming work.
Yogurt is a free, open-source TypeScript app that takes private AI meeting notes, keeping transcripts and summaries local rather than sending them to a hosted service. It drew 133 GitHub stars in its first week.
LASIK is a static CLI and MCP server that reverses Blind LLM Agents hardening so coding agents can read real code again, without executing the target. It reached 100 GitHub stars in its first week.
Brain4Agent is a Python memory module designed to give mainstream AI agents long-term memory tuned for Chinese, Japanese and Korean text; the repo drew 100 stars in its first week.
lot-map is an open-source TypeScript tool that puts a live map beside the Claude Code chat, lighting up tool calls, model steps, hooks, skills and subagents as they happen. Runs locally in five languages.
pydantic-ai is a Python framework for building AI agents, realtime voice, image generation and embeddings with typed interfaces across models. Version 2.67.0 adds SubAgents support under durable execution and input sanitization fixes.
Google released Gemini CLI 0.68.0, an open-source terminal AI agent bringing Gemini to the command line, with documentation clarifications and CLI history fixes; other CLI tools also updated.
An open-source Python project that turns a book into an Agent Skill, using a dual-engine pipeline, pluggable genre plugins and word-level fidelity verification.
ComfyUI, an open-source node-graph GUI, API and backend for running diffusion models locally, released v0.45.0 with updated inference features.
A TypeScript MCP server that gives Claude, Cursor, VS Code and Codex read-only access to Yandex Metrika analytics: reports, period comparisons, time series and exports.
An open-source TypeScript collection of Claude-style skills that run in the terminal: company-specific mock interviews, online assessments with hidden tests and resume critiques covering more than 470 companies.
OpenAI's personal finance features for ChatGPT are rolling out to the Mac desktop app, following their earlier launch elsewhere.
In iOS 27, higher iCloud+ storage tiers gain AI features for HomeKit cameras as a new perk, 9to5Mac reports.
xAI's Grok Bot is now available in Microsoft Teams, letting users bring the Grok assistant into team chats. Details on setup and features are documented on docs.x.ai.
A Google Home update brings back classic Assistant sounds for special timers, changes the smart-home confirm sound, and fixes Nest Live View issues while improving Nest Hub and Gemini for speakers.
Sorce, a Y Combinator Fall 2025 startup, lets users swipe right on jobs and sends an AI agent to submit applications. It says 850,000 people have used the tool since its August 2024 launch.
Tabletop is a Google Labs experiment where groups build interactive itineraries, polls and brackets together. Users can join a waitlist.
Firefox 160 adds on-device speech-to-text and support for MKV files with FLAC, MP3 and PCM audio, expanding browser-native media and voice capabilities.
Nameshot is a Mac app that automatically renames screenshots on-device using Apple Intelligence, so files are searchable instead of carrying timestamps. Built by an Apple Design Award-winning game designer.
Microsoft is preparing two Windows 11 changes: a unified update hub and Copilot integration in search, per NEWS.am TECH.
FourWeekMBA's ranking of 15 apps places ChatGPT first, while Muse and Claude each top a category, according to the roundup.
At Adobe MAX, an AI agent was demonstrated operating Photoshop on an Nvidia RTX Spark Surface, showing agents driving creative desktop apps end to end.
Firefox Nightly is testing Journeys, a Smart Window feature that groups related browsing history into topic-based cards with shortcuts to resume tasks.
An AI speaking partner for language learning that uses distinct personalities rather than a deferential tutor, aimed at helping learners be understood rather than fully fluent.
Valin is a private memory app that understands saved videos, posts, articles and documents, including what is said in video, so users can find things they forgot. Its AI runs entirely on-device with no account or cloud.
VoiceInk is a voice-to-text app that transcribes speech into text. Version 2.27 is out for macOS, Windows and Linux.
Typeless is an AI voice dictation app that turns speech into polished text. Version 2.11.0 is available for macOS, Windows and Linux.
Fello AI reports that ChatGPT's student discount offer ended on October 31 and rounds up what remains available for students.
More than 20,000 US veterans and transitioning service members have claimed a free year of ChatGPT Plus for career preparation.
Korean art-tech startup yourNEST launched Rheina 2.0, an emotion-aware AI recommendation engine, plus the MeM emotion-recording service and an art-commerce store, guiding users from mood input to artwork and art-tour recommendations.
Zum launched Zoe, a conversational AI assistant embedded in its CMX platform that lets district operators trace late routes, coverage gaps and pay anomalies in plain language.
At Adobe Max 2026, Adobe announced AI agents and tools for creatives, a CX Enterprise Coworker for GenStudio content workflows, and opened Creative Cloud apps to third-party AI assistants, with deeper Google and Microsoft integration.
The University of Michigan launched an AI app store for its community, Spaces4Learning reported. The store gives students and staff a curated place to find AI tools.
A Nippon.com survey finds generative AI use rising, highlighting the most common uses among respondents as adoption spreads across everyday tasks.
CarGurus is moving beyond vehicle listings to an AI agent that plays a larger role in the consumer journey and dealer workflow.
Granola is an AI-powered notepad for meetings that records and summarizes discussions. Versions 7.774.2 and 7.786.1 are released for macOS, Windows and Linux.
Zed is a multiplayer code editor built for speed with AI-assisted coding features. Version 1.28.1 is a patch release for macOS, Windows and Linux.
Red Hat Enterprise Linux 10.3 and 9.9 make AI-assisted diagnostics a supported production feature, adding OS integrity controls, AI monitoring dashboards, developer updates and Model Context Protocol support.
AWS describes an event-driven pipeline using CloudTrail, EventBridge, Lambda and AgentCore Identity that detects MCP server changes and pushes updated tool metadata to the AWS Agent Registry without stored secrets, across single- and multi-account setups.
Amp, a CLI-based frontier coding agent and development environment, released build 0.0.1794427236-ga7dfca, installable via npm as @ampcode/cli.
AWS's Generative AI Innovation Center published a catalog of 23 production-ready, open-source reusable assets covering RAG pipelines, document processors and agent scaffolding, with delivery results.
LLM, a CLI for running prompts against language models, ships 0.37.2 fixing two security issues in extra-openai-models.yaml handling, now managed by the llm openai endpoint add command.
Antigravity, Google's agent orchestration platform for coding, releases version 2.31.1 for macOS, Windows and Linux, adding a /swarm command that lets coding agents debate on a shared board.
AWS Emulator is a downloadable local emulator for AWS services. AWS says thousands of autonomous AI agents wrote every line of its code, offering a test case for agent-written software.
Google released the Gemini Decisions API, powered by its Gemini Decision model, for fast, calibrated multimodal decisions and structured outputs.
Pomerado, a YC F26 startup, converts any website into an MCP endpoint so AI agents can call an API instead of clicking through pages.
Salesforce introduced SFR-AutoR&D on November 11, 2026. Its agents produced the SFR-Embedding v4 model in about 80 hours instead of the several weeks human experimentation often requires, by autonomously changing codebases and running the research cycle.
Legal software provider LawVu said five legal AI tools — Anthropic's Claude, Microsoft Copilot, OpenAI's ChatGPT, Harvey and Legora — can now connect to its LegalOS platform with no custom development.
AWS explains how to build an AI-driven product development lifecycle with six AWS Quick agents that move from scattered customer signals to a validated prototype across Discover, Decide, Design and Prototype, with human approval at each gate.
Klad is a source-available paint app designed so a coding agent can operate it, letting agents create and manipulate images programmatically.
As AI agents become the main enterprise workload, Google's agent sandboxing aims to give each agent hard walls without runaway compute costs while they sit idle.
Databricks launched Genie Ontology and a revamped Genie, giving AI agents business context across more data sources. The company claims 1.6x accuracy using less than half the tokens.
Atuin Threads is a tool that collects AI agent sessions from every machine so you can view and resume them in one place.
A2UI 1.0 is an open protocol for generative UI that lets AI agents render interactive native screens across web, mobile and desktop, with open-source renderer SDKs for React, Angular, Lit, Flutter, SwiftUI and Jetpack Compose and agent SDKs.
Plurai raised $13M in seed funding and opened a platform that simulates and crash-tests AI agents before they reach customers, using self-driving-style end-to-end training.
Google open-sourced XSE, a system that removes Python from the serving path to speed up AI workloads, handling about six million queries a second across Search, Ads and YouTube Shorts.
A small Windows taskbar widget built with Claude Code shows Claude's 5-hour and weekly usage next to the clock, turning orange at 80% and red at 95%, with reset times on hover.
Amp is a frontier coding agent and development environment; its CLI (npm i -g @ampcode/cli) gets a new 0.0.1794424433 build with routine updates.
Alconost launched Localize.io, a localization service that uses AI agents for translation and localization workflows while keeping human review in the loop. It targets teams wanting automation without losing quality control.
A library that stores AI skills in one place and syncs them to Claude Code, Cursor and Codex, following GitHub sources, reviewing diffs and scanning each version for prompt injection, plus trimming noisy command output.
Relay lets you share a deck or doc along with the context behind it, so the recipient's AI can understand your decisions and open questions, and routes choices needing your input back to you.