Why it matters: Puts an avatar-video studio inside ChatGPT's distribution, widening AI video creation to mainstream users.
Synthesia's AI video creation and automation tools are now available inside ChatGPT, letting users generate avatar videos and automate video workflows without leaving the chatbot.
Why it matters: A 7GW compute target shows how much capital and power frontier labs need to keep scaling.
Anthropic aims to double computing capacity to 7 gigawatts next year, per an NYT report. Other reports said Amazon owns a $6.2 billion stake and that Anthropic swung to an operating loss as a large compute bill looms over its IPO.
PagerDuty launched Paige, an AI agent for site reliability engineering that works alongside ops and engineering teams, and renamed its flagship product the PD Reliability Platform, adding PD Labs.
The integration passes renters' budget, move-in date, pets and preferred amenities from Apartment List searches into leasing calls and tours, giving agents context before they speak with prospects.
Why it matters: Brings AI song generation into mainstream assistants through an MCP integration.
Suno launched an MCP beta that connects its music generation to Claude and ChatGPT, letting users create songs through those assistants via the Model Context Protocol.
Convoso released a voice AI agent that automates lead qualification and call handling for sales teams, scaling outbound calls and screening prospects without human agents.
Sanjeev Jasani and Shubham Saurabh launched TurboAgents, an AI marketing platform, ET BrandEquity reports.
ServiceNow launched Arc, an expansion of its Nvidia AI governance offering, scanx.trade reports.
Tencent Hunyuan unveiled Hyra 2.0, an AI research agent that can build its own tools and fully participate in scientific large-model and hardcore R&D work.
Why it matters: Payments infrastructure vendors are building rails specifically for AI agents that transact.
IXOPAY introduced IXO One, a modular AI-native payments platform built for agentic commerce, aimed at letting AI agents transact across payment methods and providers.
Why it matters: A huge AI-inflected buyout that puts OpenAI-affiliated capital into home security.
Thrive Holdings, backed by OpenAI, is acquiring security firm ADT in a $15.3 billion deal aimed at building AI-powered home security.
Mago's open-source AI video studio runs on ComfyUI with H3 VFX LoRAs, Qwen and Flux, guiding long-form work from script through cast, shot list and storyboards.
StackRails gives coding agents an architecture graph and policies before they write code, aiming to keep multi-repo changes consistent.
Teamily AI is an AI-native messenger where people and always-on agents work together in one context-aware space, covering docs, meetings, calls and workflows.
MachineMail gives AI agents real email inboxes to send and receive messages, connected via MCP or API, with custom domains and a trust report on inbound mail. Free tier includes 3,000 messages/month; paid from $9/month.
Hookdeck Inspect is a free webhook testing and debugging tool offering a live URL usable by developers and their agents, plus mock webhooks for 500+ providers, with no account or setup required.
Prism is an AI platform for hardware engineering that combines CAD, physics solvers, 3D viewers, coding and reporting in one agent-driven workspace. Free to try on Windows, Mac and Linux.
Synter Canvas is an AI agent that builds, launches and optimizes ad campaigns across 30+ ad platforms, with safety guardrails on every change. Charges $0.25 per verified write and 0% of ad spend.
IQ shows usage, spend and every request in one dashboard, with per-app or per-teammate keys and limits. Team plans add budgets, alerts and audit logs; built-in routing cuts costs. Free to start.
Flockbay is a free AI game maker built on Godot: an agent in the official editor writes GDScript and scenes, then playtests them by running the game and reading the debugger. Supports hosted or bring-your-own models.
Wavefunction is a minimal, free and open-source desktop coding agent for Mac that runs several tasks in parallel while you chat. Works with ChatGPT, Anthropic, Google, OpenRouter or local models, with computer use on a separate desktop.
Nana is a native Windows terminal multiplexer that runs Claude Code, Codex and other coding agents side by side, each on its own git worktree, with status and diff review.
Retailer Zalando launched an AI shopping agent to help customers find and buy products, as the company faces pressure on its margins, CNA reported.
Kravos AI launched the public beta of its enterprise-grade infrastructure platform for running AI agents.
Travel platform Klook launched K.ai, an AI agent that manages merchants' back-office operations.
Otto introduced a voice AI CRM that turns phone conversations into structured sales updates, aiming to cut manual data entry for sales teams.
Why it matters: Expands SenseTime's multimodal agent lineup into office and design workflows.
SenseTime released SenseNova 6.8 Flash, a multimodal agent model for office collaboration and design, supporting animated presentations, posters, websites, SVG graphics and 3D creation via tools like Blender.
Why it matters: Signals a major lab sees self-improving AI as an imminent milestone with safety stakes.
Anthropic says Claude now writes over 90% of its code and warns AI could soon improve itself at a dramatically accelerated pace, calling it a safety risk. The company reportedly posted a $17B quarter against a large compute bill.
Weverse Brain unveiled Agent Max, an AI agent it says drives human behavior, aimed at influencing and guiding user actions.
Redmondmag reports Microsoft is making the AI agent infrastructure that powers Copilot available to developers through Azure AI Foundry, alongside Microsoft Research demos of agentic tools for project ops, IT service desks, cloud migration and business workflows.
Why it matters: Signals how Wall Street expects AI infrastructure spending and dealmaking to be financed.
Blackstone's president and Goldman's Solomon say more fundraising and big-ticket tech deals are needed to sustain AI growth, with $720B in 2026 capex expected from Amazon, Microsoft, Google and Meta.
Why it matters: Highlights growing systemic risk concerns around AI financing and debt.
US regulators and Wall Street CEOs warn that AI investment risks and an AI-related debt binge could pressure European bond markets and strain fragile global markets if growth expectations disappoint.
Letterpress-1 is a beta coding agent, harness and IDE for growth engineering: it reads a git-based site and its market, then opens pull requests for SEO, GEO and conversion work.
SuperIT AI Agents are autonomous IT support agents that diagnose and fix incidents without runbooks. They launch in North America with a free tier for teams to try.
Runway Dev plugins feed coding agents account-aware guidance on Runway's API, so agents can call its video generation models correctly. The plugins target developers building with Runway.
Korean AI security firm S2W showcased its AI-driven security platforms at a major European cybersecurity expo.
Talkif says its voice agent platform distinguishes interruptions and caller pauses from AI decisions, enabling more precise billing for its conversational AI.
Why it matters: Signals governments expect AI-driven offensive cyber capabilities to become routine.
GCHQ's National Cyber Security Centre will warn that AI-powered attacks on UK critical infrastructure are highly likely within two years and that defences are inadequate.
The Biological Computing Co. is commercializing a video generation model derived from biological neurons, per HPCwire.
Aion released an AI agent that handles business finance work spanning banking, payments, cards and credit for companies.
AiMOGA announced MOBO, a home companion robot. The announcement was made via GlobeNewswire.
Why it matters: Simulation at this scale could cut robotaxi testing costs and timelines.
Apollo World generates up to 100,000 km of simulated driving daily, about three years of real-world testing, producing real-time multimodal camera and lidar data streams for robotaxi development.
HP launched the OmniBook 5 and ProBook 4 G2iS AI-ready laptops in India, aimed at students and mobile professionals. Pricing and availability were not detailed.
Pudu Robotics unveiled the PUDU TOMO S Series, a residential robotic mower line featuring the world's first self-sharpening rotary-blade system called SharpLoo.
Linker Vision launched Nexa, a platform combining Nvidia models, digital twins and city procedures to coordinate city responses, months after its $35M Series A.
Semiconductor IP provider Ceva has extended its NeuPro neural processing unit portfolio into AI data center infrastructure, targeting efficiency gaps rather than competing directly with Nvidia.
Nordic's nRF71 Series, announced November 3, 2026, combines Wi-Fi, Bluetooth LE, Thread, application processing and on-device AI support for battery devices, with up to 20x lower standby power than nRF70.
Oculosense launched Drishti Vision Live, described as India's first open-SDK live AI glasses, aimed at visually impaired users and industrial workers, with a developer SDK.
Why it matters: Deepens US AI chip supply chain localization for Nvidia and AMD.
Samsung plans to produce AMD and Nvidia chip components at a second Texas plant, according to a report, expanding its US advanced packaging and foundry footprint for AI chips.
Asus launched the creator-focused ProArt P14 and P16 laptops in India, powered by Nvidia RTX Spark for local AI work, starting at Rs 3,29,990 (~$3,476), with OLED displays. Separately, Microsoft and Nvidia launched a UK and Ireland AI startup program, Frontier Launchpad.
In the watchOS 27 developer beta, Apple's Audio Intelligence lets the Watch replay what someone just said and recap conversations, helping users take notes and remember audio.
D-Robotics and Unitree Robotics teamed up to launch an embodied AI education solution, pairing robot hardware with learning materials for teaching physical AI.
Why it matters: One of the largest AI infrastructure financings shows banks underwriting datacenter buildouts.
KB Kookmin Bank is lead arranger for about W12.60 trillion ($9.07B) in financing for an AI computing infrastructure project pursued by Brookfield, South Korea's largest such financing.
Mercedes-Benz launched its updated long-wheelbase E-Class in China with a China-developed AI cockpit integrating ByteDance's Doubao, priced from RMB 423,800 across three E 300 L variants.
Hua Eagle Aviation rolled out its first production HE-1A in Weinan on Oct. 29, a 4-tonne cargo drone with a 2-tonne payload. Type certification is targeted for the end of 2027.
ACTRA, the personal AI device maker, now separates its deposit from the full payment for its pendant, giving early buyers a lower entry cost. The device remains an early-bird preorder.
Why it matters: Tighter Nvidia guidance and HBM smuggling show export controls straining the AI chip supply chain.
The FT reported Nvidia ordered data center clients to reduce their China exposure, while The Information detailed smugglers moving expensive HBM memory chips from South Korea to China despite US export bans.
Krafton introduced Ludo 0.2, an AI model aimed at human-robot collaboration, according to the report.
USI launched a system-in-package module built on Wi-Fi 8 for AI devices, packing connectivity radios into a small package aimed at wearables, smart home and other edge AI hardware.
Nvidia added new settings to its Nvidia App panel for controlling DLSS graphics quality.
A UK insurance policy has launched to cover cases where a domestic humanoid robot injures a person, an early commercial product for household robot liability.
Facing labor shortages, Japan's JR East is putting a chainsaw-armed humanoid robot to work trimming trees, an example of physical AI in railway maintenance.
ACTRA has opened early-bird access in the US for a personal AI device pitched as going beyond the smartphone; hardware and pricing details were not specified.
Nothing OS 5.0, based on Android 17, adds an AI app builder and new Glyph features; Livemint details the rollout timeline and how to download it.
Vultr launched a 20 MW AI supercompute cluster in Chennai, India, expanding cloud GPU capacity for regional AI workloads.
PagerDuty launched Paige, an AI agent for site reliability engineering that works alongside ops and engineering teams, and renamed its flagship product the PD Reliability Platform, adding PD Labs.
Typeless is AI voice dictation software that turns speech into cleaned-up, polished text. Version 2.10.0 is out for macOS, Windows and Linux.
Lovable co-founder and CTO Fabian Hedin spoke at Bloomberg Tech 2026 about how AI is changing software creation, the engineer's role and the challenge of building secure, maintainable apps.
Anthropic added a new rule to Claude Code stating that using Claude to infringe intellectual property, such as cloning games, is against its policies.
Open WebUI is a user-friendly self-hosted interface for running local and API models (Ollama, OpenAI). Version 0.13.0 adds a tabbed workspace for files, terminal sessions, artifacts, previews, voice calls and comments.
Analysis of how agentic AI could enable rapid low-code modernization of COBOL and mainframe systems, and what that means for enterprise IT's long-standing workhorse.
Google Home's latest updates include smoother Gemini voice responses, Nest Cam improvements and better Matter support, T3 reports.
A Hugging Face GGUF of DeepSeek-V4.1-Flash-1017 compressed to about 1 bit per parameter, with half its 384 experts per layer removed, yielding 72.1 GB of weights for code and agent workflows.
ByteDance's Doubao chatbot has reached 419 million users, with a daily base roughly six times that of Alibaba's Qwen, per FourWeekMBA.
Vivo is rolling out OriginOS 7, based on Android 17, to X300, X300 Pro and X300 FE, bringing more powerful AI features and stronger cross-ecosystem connectivity.
A Washington Post report describes a school using AI to generate a 'future self' of a six-year-old student based on her imagined career, raising questions about AI in education.
Ollama, a tool for running large language models locally on your own computer, released version 0.40.14 for macOS, Windows and Linux.
StackRails gives coding agents an architecture graph and policies before they write code, aiming to keep multi-repo changes consistent.
Amp's CLI, a command-line interface for the Amp frontier coding agent and development environment, gets a new build installable via npm with @ampcode/cli.
Meta's AI-powered creator assistant adds proactive trend alerts, better content planning and automated replies to help creators perform across Facebook and Instagram.
OpenAI's ChatGPT Dots upgrade so the feature selects the appropriate model for each task automatically, without user input. Users also report new usage limits that some subscribers criticize.
Teamily AI is an AI-native messenger where people and always-on agents work together in one context-aware space, covering docs, meetings, calls and workflows.
MachineMail gives AI agents real email inboxes to send and receive messages, connected via MCP or API, with custom domains and a trust report on inbound mail. Free tier includes 3,000 messages/month; paid from $9/month.
Hookdeck Inspect is a free webhook testing and debugging tool offering a live URL usable by developers and their agents, plus mock webhooks for 500+ providers, with no account or setup required.
Prism is an AI platform for hardware engineering that combines CAD, physics solvers, 3D viewers, coding and reporting in one agent-driven workspace. Free to try on Windows, Mac and Linux.
IQ shows usage, spend and every request in one dashboard, with per-app or per-teammate keys and limits. Team plans add budgets, alerts and audit logs; built-in routing cuts costs. Free to start.
Flockbay is a free AI game maker built on Godot: an agent in the official editor writes GDScript and scenes, then playtests them by running the game and reading the debugger. Supports hosted or bring-your-own models.
Wavefunction is a minimal, free and open-source desktop coding agent for Mac that runs several tasks in parallel while you chat. Works with ChatGPT, Anthropic, Google, OpenRouter or local models, with computer use on a separate desktop.
Nana is a native Windows terminal multiplexer that runs Claude Code, Codex and other coding agents side by side, each on its own git worktree, with status and diff review.
Version 3.28.7 of Cursor, the AI code editor for writing, editing and chatting about code, is out for macOS, Windows and Linux.
Kravos AI launched the public beta of its enterprise-grade infrastructure platform for running AI agents.
An open GitHub project that lets AI assistants such as ChatGPT and Claude remotely control a computer through MCP, with security controls. It drew 102 stars in its first week.
A dataset of 4,146 Three.js 3D scene examples with prompts, thinking traces and full HTML/JavaScript responses, generated with MiniMax M3.1 Flash. Covers objects, interiors, architecture and voxel scenes.
Balanov/skein is a Go project for orchestration of agent swarms that fails closed, using epoch-fenced work units and rejection of late completions. It drew 116 stars in its first week.
Weverse Brain unveiled Agent Max, an AI agent it says drives human behavior, aimed at influencing and guiding user actions.
A Hugging Face GGUF of DeepSeek-V4.1-Flash-1017 compressed to about 1 bit per parameter, with half its 384 experts per layer removed, yielding 72.1 GB of weights for code and agent workflows.
A dataset of 4,146 Three.js 3D scene examples with prompts, thinking traces and full HTML/JavaScript responses, generated with MiniMax M3.1 Flash. Covers objects, interiors, architecture and voxel scenes.
A Hugging Face Space runs FastH3-Omni, a distilled student of MiniMax-H3 that turns ordered image, video and audio references plus a prompt into a short clip with matching soundtrack in eight steps.
A Hugging Face repository, ML-Intern-lab/Qwen3.5-0.8B-weight-space-programming, offers an image-text-to-text model for reading and writing neural network weights. It is a public repo, not an inference endpoint.
A paper proposes Reviso, an agentic video captioning framework arguing that not all visual evidence deserves equal attention, so MLLMs should weight key events over minor details.
CoworkOS is a research system for general-purpose agent coworkers that must deliver complex productivity work while monitoring and deciding over long horizons.
Phonon is an open text-to-speech model family; this Hugging Face Space demo runs it entirely in your browser in English, French, German, Spanish and Portuguese, with no audio leaving the device.
OBLITERATUS published DeepSeek-V4.1-Flash-1017-obliterated, a refusal-direction ablated version of DeepSeek V4.1 Flash with BF16 weights under MIT, reporting a 4% refusal rate.
Open WebUI is a user-friendly self-hosted interface for running local and API models (Ollama, OpenAI). Version 0.13.0 adds a tabbed workspace for files, terminal sessions, artifacts, previews, voice calls and comments.
An open GitHub project that lets AI assistants such as ChatGPT and Claude remotely control a computer through MCP, with security controls. It drew 102 stars in its first week.
Balanov/skein is a Go project for orchestration of agent swarms that fails closed, using epoch-fenced work units and rejection of late completions. It drew 116 stars in its first week.
Pydantic-ai, a Python library for building typed AI agents with voice, image generation and embeddings, released v2.65.0 with compatibility notes, multimodal AskUserAnswer additions and browser-use call counting.
A Python repo of 32 Agent Skills for Claude Code and Codex aimed at Chinese doctors and medical students, covering topic selection, PubMed search, bilingual writing, journal submission and peer-review responses.
A TypeScript MCP server and CLI that finds Brazilian companies by CNAE code and region using open CNPJ data from Receita Federal.
An open-source TypeScript project that lets a company set up Claude Code and Codex once, self-hosted, with shared company files, team skills, MCP and per-person folder permissions.
dsh-see-now is a JavaScript tool that captures your screen, filters frames by how much they change and tiles several into one image, so AI can watch your screen in real time without spending tokens on static content.
JavaScript GitHub project adding a memory layer to the AstrBot chatbot: five memory types, hybrid retrieval, forgetting rules, emotional boundaries and an autonomous life loop, without altering personality.
JavaScript GitHub project for running language-model agents inside Minecraft with Mineflayer tools, persistent sessions and shell sandboxes.
NextTransformer is an open-source Python framework for next-generation Transformer architectures, offering compute-matched benchmarking and a platform for autonomous architecture research. It drew 101 stars in its first week.
FlashAlign is a Python tool that re-aligns AI models instantly, either per request for verified users and channels or always on, and works with Hermes, OpenClaw and custom harnesses. It reached 100 stars in its first week.
omux is a native macOS terminal app for running many Claude Code and Codex sessions side by side, with projects, worktrees, live agent status, sidebar allow/deny, resumable conversations and PR status.
uncut is a 7 MB coding-agent harness for shell, CI and local models that claims lossless context with no silent truncation and no TUI. Written in Go, it gained 121 GitHub stars in its first week.
AnyModelBot is an open-source, local-first alternative to hosted AI bots that runs agents on 70+ models via one Modelflare key, reachable in-app, WeChat or Feishu, with memory and computer control.
graphmemory is an open-source memory system for AI apps: it stores facts bitemporally, applies time-graded recall and decay, expires entries by TTL and dedupes, backed by Neo4j, PostgreSQL and Redis.
Typeless is AI voice dictation software that turns speech into cleaned-up, polished text. Version 2.10.0 is out for macOS, Windows and Linux.
Google Home's latest updates include smoother Gemini voice responses, Nest Cam improvements and better Matter support, T3 reports.
ByteDance's Doubao chatbot has reached 419 million users, with a daily base roughly six times that of Alibaba's Qwen, per FourWeekMBA.
Vivo is rolling out OriginOS 7, based on Android 17, to X300, X300 Pro and X300 FE, bringing more powerful AI features and stronger cross-ecosystem connectivity.
A Washington Post report describes a school using AI to generate a 'future self' of a six-year-old student based on her imagined career, raising questions about AI in education.
Meta's AI-powered creator assistant adds proactive trend alerts, better content planning and automated replies to help creators perform across Facebook and Instagram.
OpenAI's ChatGPT Dots upgrade so the feature selects the appropriate model for each task automatically, without user input. Users also report new usage limits that some subscribers criticize.
Teamily AI is an AI-native messenger where people and always-on agents work together in one context-aware space, covering docs, meetings, calls and workflows.
MachineMail gives AI agents real email inboxes to send and receive messages, connected via MCP or API, with custom domains and a trust report on inbound mail. Free tier includes 3,000 messages/month; paid from $9/month.
Weverse Brain unveiled Agent Max, an AI agent it says drives human behavior, aimed at influencing and guiding user actions.
TeslaNorth reports Grok Bot's latest update adds shopping and payments capabilities plus support for team bots, part of nine upgrades to the bot product.
Genspark, an all-in-one AI workspace whose Super Agent handles research and tasks from a prompt, shipped an iOS update fixing known issues.
Contentsquare released an AI-first experience platform that optimizes both human and AI-agent customer journeys, aimed at analytics and personalization across digital channels.
OpenAI is retiring the ChatGPT Classic app in favor of its desktop app; the Classic app will shut down on January 30, 2027.
Inspired by Andrej Karpathy's Claude-built HackerTube, a YouTube clone showing only Hacker News-suggested videos, Own Your HackerTube lets anyone create their own personal AI-filtered version.
ON1 released Photo RAW 2027 for Mac and Windows, adding a new RAW engine and revised NoNoise AI models. The AI Culling Assistant is exclusive to the MAX edition.
Readwise Books combines an ebook store, reading tracker, detailed reading stats and AI features, aiming to challenge Amazon's reading ecosystem.
Granola is an AI-powered notepad that listens to meetings and turns transcripts into structured notes. Version 7.726.2 is a minor app update with no details in the source.
Moonshot's Kimi AI chat assistant released version 3.2.24 for macOS, Windows and Linux. Separately, a tutorial walks through building an agent on Kimi K3.1 Pro in 13 steps.
A new Chrome AI feature uses Gemini to read Gmail and autofill web forms; Google may already be reconsidering the change, and users can opt out.
BoltAI is a desktop AI chat client that lets you use multiple AI models from one app. Version 2.18.4 is out for macOS, Windows and Linux.
warrantini launched a way to run warranty operations inside Claude, ChatGPT and Microsoft Copilot, letting assistants handle warranty-related tasks.
A report says being rude to a particular AI chatbot could get users banned, explaining the reason behind the policy.
Drive Tesla Canada reports Tesla could soon add xAI's Grok assistant to its Intel-based vehicles, expanding the chatbot beyond newer hardware.
ElevenLabs' ElevenReader app turns books, PDFs, docs and articles into natural-sounding audio. Version 1.11.40 for iPhone/iPad adds suggested Collections, a true black reading background and in-app problem reporting.
PagerDuty launched Paige, an AI agent for site reliability engineering that works alongside ops and engineering teams, and renamed its flagship product the PD Reliability Platform, adding PD Labs.
Lovable co-founder and CTO Fabian Hedin spoke at Bloomberg Tech 2026 about how AI is changing software creation, the engineer's role and the challenge of building secure, maintainable apps.
Anthropic added a new rule to Claude Code stating that using Claude to infringe intellectual property, such as cloning games, is against its policies.
Analysis of how agentic AI could enable rapid low-code modernization of COBOL and mainframe systems, and what that means for enterprise IT's long-standing workhorse.
Ollama, a tool for running large language models locally on your own computer, released version 0.40.14 for macOS, Windows and Linux.
StackRails gives coding agents an architecture graph and policies before they write code, aiming to keep multi-repo changes consistent.
Amp's CLI, a command-line interface for the Amp frontier coding agent and development environment, gets a new build installable via npm with @ampcode/cli.
Hookdeck Inspect is a free webhook testing and debugging tool offering a live URL usable by developers and their agents, plus mock webhooks for 500+ providers, with no account or setup required.
Prism is an AI platform for hardware engineering that combines CAD, physics solvers, 3D viewers, coding and reporting in one agent-driven workspace. Free to try on Windows, Mac and Linux.
IQ shows usage, spend and every request in one dashboard, with per-app or per-teammate keys and limits. Team plans add budgets, alerts and audit logs; built-in routing cuts costs. Free to start.
Flockbay is a free AI game maker built on Godot: an agent in the official editor writes GDScript and scenes, then playtests them by running the game and reading the debugger. Supports hosted or bring-your-own models.
Wavefunction is a minimal, free and open-source desktop coding agent for Mac that runs several tasks in parallel while you chat. Works with ChatGPT, Anthropic, Google, OpenRouter or local models, with computer use on a separate desktop.
Nana is a native Windows terminal multiplexer that runs Claude Code, Codex and other coding agents side by side, each on its own git worktree, with status and diff review.
Version 3.28.7 of Cursor, the AI code editor for writing, editing and chatting about code, is out for macOS, Windows and Linux.
Kravos AI launched the public beta of its enterprise-grade infrastructure platform for running AI agents.
Redmondmag reports Microsoft is making the AI agent infrastructure that powers Copilot available to developers through Azure AI Foundry, alongside Microsoft Research demos of agentic tools for project ops, IT service desks, cloud migration and business workflows.
Letterpress-1 is a beta coding agent, harness and IDE for growth engineering: it reads a git-based site and its market, then opens pull requests for SEO, GEO and conversion work.
Microsoft Research write-up on assessing SQL estates ahead of Fabric and AI migration. No tool release, pricing or availability details in the source.
SuperIT AI Agents are autonomous IT support agents that diagnose and fix incidents without runbooks. They launch in North America with a free tier for teams to try.
Version 1.144.1 of Microsoft's open-source code editor is out for macOS, Windows and Linux.
Claude Code 2.1.317 is out for CLI and VS Code, letting Claude edit files, run terminal commands and handle workflows from the terminal. Install via npm.
LLM, a CLI tool and Python library for running prompts against language models, hits 0.37 with gpt-6.1-sol and gpt-6.1-terra support plus an llm openai endpoint add command for configuring extra models.
Latent Space interviews Legora CTO Jacob Lauritzen on building and testing legal AI agents when the firm cannot see what the agent does for a lawyer because data belongs to law firm clients.
Moss-Image turns JSON prompts into pixel-perfect layouts; a community layout editor lets users build those prompts and pair them with an LLM to generate designs.
Dyad is an AI-powered app builder. Version 1.22.0 is out for macOS, Windows and Linux.