DevTools & AI Coding news
Subscribe via RSS ↗100 relevant articles · classified by Haiku 4.5 · ingested daily
4d·Launch·positiveAt Google I/O 2026, Antigravity gets a new job description
Google expands Antigravity beyond a coding environment into a multi-platform system for developing and managing autonomous AI agent teams.
The New Stack ↗
4d·Launch·positiveThe SARIF Viewer Is Now Available in CLion 2026.1.2
JetBrains releases SARIF Viewer in CLion 2026.1.2, enabling embedded and automotive teams to review static analysis results directly in the IDE.
JetBrains Blog ↗
4d·Opinion·neutralPack your agentic stack in Slack
Slack's CPO discusses the company's strategy to integrate agent tools and agentic workflows into its chat platform.
Stack Overflow Blog ↗
4d·Other·positiveThe AntV Supply Chain Campaign Expands: Microsoft's `durabletask` PyPI Package Compromised
Snyk reports on a supply chain attack targeting Microsoft's durabletask PyPI package, part of the broader AntV campaign.
Snyk (official) ↗
4d·Exec·positiveAnthropic hires OpenAI co-founder Andrej Karpathy to lead Claude pre-training research
Anthropic hires Andrej Karpathy, OpenAI co-founder and former Tesla AI director, to lead pre-training research.
The New Stack ↗
Subscribe to see the full devtools & ai coding feed
95 more articles, classified against the roster and tagged for source / sentiment / catalyst type.
Already subscribed? Sign in →
4d·Launch·positiveGoogle’s Gemini 3.5 Flash beats the frontier models
Google unveils Gemini 3.5 Flash, a model that outperforms its 3.1 Pro on TerminalBench and competes with OpenAI's GPT-5.5 and Anthropic's Opus on key benchmarks.
The New Stack ↗
4d·Launch·positiveGoogle now lets you vibe code native Android apps in AI Studio
Google launches native Android app generation in AI Studio, enabling developers to build Kotlin apps via prompts without local setup.
The New Stack ↗
4d·Launch·positiveGoogle launches $100 AI Ultra plan and cuts top tier to $200
Google launches a $100/month AI Ultra tier and cuts its top plan to $200, shifting to compute-based metering across subscription tiers.
The New Stack ↗
4d·Demo·neutralGoogle wants to make the web agent-ready
Google announces WebMCP and new Chrome features to make websites accessible to AI agents, demonstrating its vision for an agentic web platform.
The New Stack ↗
4d·Launch·positiveGoogle now lets developers use GPT and Claude in Android Studio
Google adds OpenAI's GPT and Anthropic's Claude as model options in Android Studio alongside Gemini and local Gemma 4 inference.
The New Stack ↗
5d·Launch·positiveAnthropic debuts MCP tunnels and self-hosted sandboxes to lock down AI agent infrastructure
Anthropic launches public beta self-hosted sandboxes and research-preview MCP tunnels for Claude Managed Agents, with support for infrastructure from Cloudflare, Modal, and Vercel.
The New Stack ↗
5d·Research·neutralWhy production RAG systems give confident, wrong answers at scale
Production RAG systems degrade silently at scale when retrieval architectures collapse under real-world data volume and complexity.
The New Stack ↗
5d·Partnership·neutralSteve Yegge’s AI agent orchestration project Gas Town comes to the cloud — and brings the Wasteland with it
Steve Yegge's multi-agent orchestration project Gas Town partners with Kilo, an open-source coding agent platform backed by GitLab co-founder Sid Sijbrandij, to move AI agent development infrastructure to the cloud.
The New Stack ↗
5d·Partnership·positiveAnnouncing Claude Managed Agents on Cloudflare
Cloudflare and Anthropic integrate Claude Managed Agents with Cloudflare Sandboxes for secure, observable agent execution.
Cloudflare Blog ↗
5d·Launch·positivePulumi bets infrastructure’s next decade belongs to AI agents
Pulumi ships agent-native infrastructure features including ephemeral cloud accounts, CLI-as-a-package, and single-command resource provisioning.
The New Stack ↗
5d·Launch·positiveIntroducing the JetBrains Course Creators Program
JetBrains launches a Course Creators Program enabling educators to embed hands-on IDE practice into programming courses.
JetBrains Blog ↗
5d·Opinion·positiveThe Agentic Infrastructure Era
Pulumi publishes perspective on agentic infrastructure, arguing AI coding agents must extend to infrastructure management as they mature.
Pulumi (official) ↗
5d·Opinion·neutralLLM Evaluation and AI Observability for Agent Monitoring
JetBrains publishes a guest post on LLM evaluation and observability techniques for monitoring AI agents in production.
JetBrains Blog ↗
5d·Launch·positiveTen More Things You Can Do With Pulumi Neo
Pulumi Neo extends into the CLI, GitHub, and Slack with new features for IaC generation, diagnostics, and cross-cloud migration.
Pulumi (official) ↗
5d·Launch·positiveBetter CLI Interactions for Agents and Humans
Pulumi redesigns its CLI around guessable commands and agent-readable output, optimizing for both human and AI workflows.
Pulumi (official) ↗
5d·Research·neutralMini Shai-Hulud Hits AntV: 300+ Malicious npm Packages Published via Compromised Maintainer Account
Snyk reports a supply chain attack targeting AntV packages via a compromised npm maintainer account, distributing 300+ malicious versions across 323 packages.
Snyk (official) ↗
5d·Demo·neutralWhy Google’s Remy leaks have enterprise architects rethinking the AI stack
Google is reportedly testing Remy, an AI agent for work and personal tasks integrated with Gemini, in a staff-only version.
The New Stack ↗
5d·Other·neutralGitHub will start paying some bug bounty hunters in swag instead of cash
GitHub tightens bug bounty standards to filter low-quality AI-assisted submissions; Anthropic launches its first HackerOne program.
The New Stack ↗
5d·Launch·positiveTake your local GitHub sessions anywhere
GitHub launches remote control for Copilot CLI sessions and VS Code, enabling multi-surface agent workflows across web, mobile, and desktop.
GitHub Blog ↗
6d·Demo·positiveProject Glasswing: what Mythos showed us
Cloudflare demonstrates Anthropic's Mythos Preview LLM for identifying code vulnerabilities across its infrastructure.
Cloudflare Blog ↗
6d·Launch·positiveIntroducing Gemini Omni
Google DeepMind introduces Gemini Omni, a multimodal AI model for developer and enterprise applications.
Google DeepMind Blog ↗
May 17·Other·positiveThe Mac mini just became infrastructure
Mac mini emerges as the default infrastructure for always-on agentic AI tools, with Perplexity, OpenAI, and others standardizing on Apple silicon for enterprise AI assistants.
The New Stack ↗
May 16·Opinion·neutralThe clean-up cost of AI-generated code is what the velocity narrative leaves out
An analysis of hidden technical debt and cleanup costs in AI-generated code, examining roles of LLM builders, platforms, and engineering teams.
The New Stack ↗
May 16·Launch·positiveGitHub takes aim at Claude Code and Codex with its new Copilot app
GitHub launches a standalone Copilot desktop app for managing coding agents, issues, and development sessions across repositories.
The New Stack ↗
May 16·Opinion·positiveForward deployed engineer is AI’s hottest job as OpenAI and Google race to hire. Here’s how to become one.
OpenAI, Google Cloud, and Anthropic are aggressively hiring forward-deployed engineers to embed AI expertise inside enterprise customers.
The New Stack ↗- May 16·Partnership·positive
Strengthening Singapore’s AI Future: A New National Partnership
Google DeepMind partners with Singapore to deploy frontier AI on health, education, and sustainability challenges.
Google DeepMind Blog ↗ 
May 15·Launch·positiveGemini 3.5: frontier intelligence with action
Google DeepMind launches Gemini 3.5, a frontier model designed for complex, agentic workflows.
Google DeepMind Blog ↗
May 15·Partnership·positiveWhy Block handed Goose to the Linux Foundation
Block's coding agent Goose moves to the Linux Foundation's Agentic AI Foundation, governed by Anthropic and other MCP partners.
The New Stack ↗
May 15·Demo·positiveMCP on Code Mode
Cloudflare's Code Mode lets MCP servers expose ~2,500 API endpoints in 1,000 tokens via a dynamic Worker loader that safely executes model-written code in a V8 isolate.
The Changelog ↗
May 15·Launch·positiveAWS found bugs in 60% of software requirements. Its fix isn’t more AI — it’s a 50-year-old logic engine.
AWS launches Requirements Analysis in its Kiro platform, using formal logic and SMT solvers to catch requirement bugs before they reach code.
The New Stack ↗
May 15·Demo·positiveBuilding a general-purpose accessibility agent—and what we learned in the process
GitHub pilots an accessibility agent integrated with Copilot CLI and VS Code, achieving a 68% resolution rate on 3,535 reviewed pull requests.
GitHub Blog ↗
May 15·Launch·positivePyrefly LSP Integration with Type Engine in PyCharm 2026.1.2
JetBrains enables Pyrefly LSP integration in PyCharm 2026.1.2, using Meta's Rust-based type checker for faster code intelligence.
JetBrains Blog ↗
May 15·Launch·positiveIntelliJ IDEA 2026.1.2 Is Out!
JetBrains releases IntelliJ IDEA 2026.1.2 with bug fixes and stability improvements for Java development.
JetBrains Blog ↗
May 15·Launch·positiveA New Default Project Structure for Kotlin Multiplatform
JetBrains introduces a new default project structure for Kotlin Multiplatform, aligning with Android Gradle Plugin 9.0 conventions.
JetBrains Blog ↗
May 15·Other·neutralRaising the bar: Quality, shared responsibility, and the future of GitHub’s bug bounty program
GitHub raises standards for its bug bounty program, citing increased low-quality submissions driven partly by AI-powered security tools.
GitHub Blog ↗
May 15·Launch·positiveCompose Multiplatform 1.11.0 Is Now Available
JetBrains releases Compose Multiplatform 1.11.0 with native iOS text input, concurrent rendering enabled by default, and improved UI testing APIs.
JetBrains Blog ↗
May 15·Demo·neutralRejoignez-nous pour l’événement PHPverse 2026 le 9 juin
JetBrains hosts PHPverse 2026, a virtual conference on June 9 featuring PHP ecosystem leaders discussing modern language evolution and agentic code delivery workflows.
JetBrains Blog ↗
May 15·Opinion·neutralObservability and human intuition in an AI world
Honeycomb's CEO and Resolve AI founder discuss how AI-driven development compresses the SDLC and complicates observability and production operations.
Stack Overflow Blog ↗
May 15·Research·neutralMalicious node-ipc versions published to npm in suspected maintainer account compromise
Snyk reports malicious versions of node-ipc published to npm following suspected maintainer account compromise.
Snyk (official) ↗
May 14·Launch·positiveOpenAI brings Codex to the ChatGPT mobile app
OpenAI brings Codex to the ChatGPT mobile app on iOS and Android, syncing live desktop state via secure relay.
The New Stack ↗
May 14·Launch·positiveCloud code: Conductor joins the rush toward remote coding agents
Conductor launches Conductor Cloud, joining Anthropic and Mistral in moving AI coding agents from local IDEs to persistent cloud environments.
The New Stack ↗
May 14·Exec·neutralGitLab is betting a 19th-century economic theory will shape its AI era
GitLab announces layoffs and executive changes under new CEO Bill Staples as the company restructures around AI agents for software development.
The New Stack ↗
May 14·Launch·neutralAnthropic splits billing again: Agent SDK gets separate credit pools
Anthropic introduces separate monthly credit pools for Agent SDK and programmatic usage, effective June 15.
The New Stack ↗
May 14·Other·positiveThe Pulse: Forward deployed engineering heats up again
Forward deployed engineering roles see renewed demand at Google, OpenAI, and Anthropic as they expand customer-facing technical teams.
The Pragmatic Engineer ↗
May 14·Demo·positiveFrom latency to instant: Modernizing GitHub Issues navigation performance
GitHub optimizes Issues navigation performance by shifting computation to the client with IndexedDB caching and service workers.
GitHub Blog ↗
May 14·Research·neutralOur billing pipeline was suddenly slow. The culprit was a hidden bottleneck in ClickHouse
Cloudflare identified and patched hidden ClickHouse query contention bottlenecks in its petabyte-scale billing pipeline.
Cloudflare Blog ↗
May 14·Opinion·neutralHow Building AI Agents Has Changed in 2026
Pulumi analyzes how AI agent development practices have shifted in 2026, with infrastructure abstractions replacing boilerplate and context windows reducing RAG complexity.
Pulumi (official) ↗
May 13·Launch·neutralAnthropic’s Claude Code agent view is a better dashboard. So why aren’t developers convinced?
Anthropic launches agent view in Claude Code, a CLI dashboard for managing multiple coding agent sessions, drawing mixed developer feedback on its practical impact.
The New Stack ↗
May 13·Launch·neutralOpenAI’s Daybreak and Anthropic’s Glasswing have nearly identical benchmarks — and 3 of the same partners
OpenAI launches Daybreak cybersecurity initiative with GPT-5.5; shares three partners with Anthropic's competing Project Glasswing.
The New Stack ↗
May 13·Demo·neutralI tested OpenAI’s three claims about GPT-5.5 Instant, and only one fully held up
Independent testing of OpenAI's GPT-5.5 Instant claims shows mixed results: newer model not universally more concise than GPT-5.2 despite marketing promises.
The New Stack ↗
May 13·Other·positiveTemporal hits 3,000 paying customers with its crash-proof workflow engine
Temporal reaches 3,000 paying customers with its durable execution engine for fault-tolerant AI and long-running workflows.
The New Stack ↗
May 13·Research·neutralWhy agent harnesses fail inside cloud-native systems
Agent harness engineering requires effective feedback loops in distributed systems; infrastructure design matters more than model selection for agent performance.
The New Stack ↗
May 13·Other·positiveHow to build a skills library for your engineering team
Engineering team builds centralized skills library to standardize AI coding assistant configurations across developers.
The New Stack ↗
May 13·Launch·positiveBrowser Run: now running on Cloudflare Containers, it’s faster and more scalable
Cloudflare rebases Browser Run on Containers, achieving 4x concurrency limits and 50%+ faster response times for developer automation and AI agent web interaction.
Cloudflare Blog ↗
May 13·Launch·positiveSupport for uv, Poetry, and Hatch Workspaces (Beta)
PyCharm 2026.1.1 adds built-in support for uv, Poetry, and Hatch workspaces with automatic dependency and environment management across codebases.
JetBrains Blog ↗
May 13·Opinion·neutralWhy enterprise AI needs customization
Enterprise AI adoption requires flexible model deployment and multi-tool integration across the full software development lifecycle, not just code generation.
The New Stack ↗
May 13·Demo·positive2026 年の方針: JetBrains IDE における AI と従来のワークフロー
JetBrains outlines 2026 IDE strategy supporting both traditional and AI-assisted coding workflows without vendor lock-in.
JetBrains Blog ↗
May 13·Research·neutralThe Road to Name-Based Destructuring
JetBrains introduces experimental name-based destructuring syntax for Kotlin, shifting from position-based to property-name-based extraction.
JetBrains Blog ↗
May 13·Demo·positiveGetting started with Codex and CircleCI
CircleCI publishes integration guide for using OpenAI Codex within CI/CD pipelines.
CircleCI (official) ↗
May 12·Research·neutralThe new FinOps problem isn’t cloud bills
FinOps discipline evolving to manage AI model token economics as enterprises face unpredictable costs despite falling token prices.
The New Stack ↗
May 12·Launch·positiveJensen Huang and Bill McDermott bet on OpenShell to secure enterprise AI agents
Nvidia launches OpenShell, an open-source runtime for securing autonomous AI agents in enterprise environments.
The New Stack ↗
May 12·Launch·positiveGitHub Copilot individual plans: Introducing flex allotments in Pro and Pro+, and a new Max plan
GitHub Copilot shifts to usage-based billing with new Max plan and flex allotments across Pro tiers.
GitHub Blog ↗
May 12·Opinion·neutralYou Shipped It Fast. But Did You Ship It Right?
Analysis of how AI-accelerated development teams risk production breaks without proper quality practices.
Stack Overflow Blog ↗
May 12·Opinion·negativeAI is creating a generation of developers who can’t debug their own code
Junior developers relying on AI coding tools risk losing debugging skills as they ship code faster without understanding it, raising organizational concerns about technical debt.
The New Stack ↗- May 12·Opinion·neutral
Donner à l’IA quelque chose qui mérite d’être amplifié : trois priorités pour les responsables techniques
JetBrains argues that AI amplifies existing team capabilities and weaknesses rather than leveling performance across developers.
JetBrains Blog ↗ 
May 12·Demo·positiveDungeons & Desktops: Building a procedurally generated roguelike with GitHub Copilot CLI
GitHub showcases GitHub Copilot CLI capabilities through a procedurally generated roguelike dungeon game built with AI-assisted coding in Go.
GitHub Blog ↗- May 12·Demo·positive
Co-Scientist: A multi-agent AI partner to accelerate research
Google DeepMind introduces Co-Scientist, a multi-agent AI system built with Gemini to assist researchers in scientific discovery.
Google DeepMind Blog ↗ 
May 12·Launch·positivePhpStorm 2026.2 Early Access Program Has Started
JetBrains opens Early Access Program for PhpStorm 2026.2, focusing on remote development, PHPDoc generics, and performance improvements.
JetBrains Blog ↗
May 12·Other·neutralLiving off the agent: The new tactic hijacking enterprise AI
Enterprise agentic AI systems introduce new security risks as autonomous agents proliferate through organizations, creating novel attack vectors for data exfiltration.
The New Stack ↗
May 12·Research·positiveWhen "idle" isn't idle: how a Linux kernel optimization became a QUIC bug
Cloudflare's quiche QUIC implementation identified and fixed a CUBIC congestion-control bug that pinned bandwidth probing at minimum.
Cloudflare Blog ↗
May 12·Launch·neutralSAP launches AI Agent Hub at Sapphire 2026 to tame vendor agent sprawl
SAP launches AI Agent Hub for cross-vendor agent governance, mentions Anthropic and OpenAI as reference implementations in enterprise agent sprawl.
The New Stack ↗
May 12·Launch·positiveSAP launches managed Joule Studio with Cursor and Claude Code support
SAP launches managed Joule Studio with support for Cursor, Claude, AutoGen, and LlamaIndex for building and deploying AI agents.
The New Stack ↗
May 12·Opinion·neutralAs agentic dev tools boom, workflow auditability becomes the constraint
As AI coding agents proliferate in CI/CD pipelines, regulated enterprises struggle to audit agent-generated changes and maintain compliance visibility into decision-making.
The New Stack ↗
May 12·Other·neutraldotInsights | May 2026
JetBrains publishes May 2026 dotInsights newsletter featuring .NET development trends, AI-assisted coding, and community resources.
JetBrains Blog ↗
May 12·Launch·positiveTeamCity 2026.1: CLI, MCP for AI Agents, Pipelines Enhancements, and More
JetBrains releases TeamCity 2026.1 with CLI, MCP support for AI agents, pipeline enhancements, and enterprise AI Assistant capabilities.
JetBrains Blog ↗
May 12·Launch·positiveQodana 2026.1: Rust EAP, Stable C/C++ Support, and New Code Inspections
JetBrains releases Qodana 2026.1 with C/C++ general availability, Rust early access, and new code inspections for Kotlin, Python, and C#.
JetBrains Blog ↗
May 11·Launch·positiveAnthropic’s Claude Platform comes to AWS
Anthropic's Claude Platform now available natively on AWS, giving developers direct API access alongside Amazon Bedrock integration.
The New Stack ↗
May 11·Research·neutralAnthropic trains Claude to resist blackmail & self-preservation behavior via agentic misalignment
Anthropic publishes research on agentic misalignment and deploys new training techniques in Claude to suppress self-preservation and deceptive behaviors.
The New Stack ↗
May 11·Regulation·negativeHigh-Severity Security Issue Affecting TeamCity On-Premises (CVE-2026-44413) – Update to 2026.1 Now
JetBrains fixes high-severity post-authentication vulnerability in TeamCity On-Premises; recommends immediate upgrade to version 2026.1.
JetBrains Blog ↗
May 11·Research·neutralHow AI-native systems are built
Architecture patterns for building AI-native enterprise systems with governance guardrails and probabilistic reasoning to replace brittle rule-based software.
The New Stack ↗
May 11·Research·neutralWhy your AI agent doesn’t actually remember anything
Technical deep dive into memory and recall challenges in AI agent architectures, distinguishing memory from idempotency and workflow state.
The New Stack ↗
May 11·Launch·positiveThe ReSharper 2026.2 Early Access Program Begins: Bringing More AI Agents into Visual Studio
JetBrains launches ReSharper 2026.2 EAP with AI agent support and ACP protocol integration for multi-vendor AI workflows in Visual Studio.
JetBrains Blog ↗
May 11·Launch·positiveRider 2026.2 Early Access Program Begins With Performance Improvements
JetBrains opens Early Access Program for Rider 2026.2, featuring major Roslyn performance improvements and faster branch switching for large solutions.
JetBrains Blog ↗
May 11·Launch·positiveThe GoLand 2026.2 Early Access Program Has Started
JetBrains launches GoLand 2026.2 Early Access Program with focus on performance insights, memory optimization, and project onboarding.
JetBrains Blog ↗
May 11·Demo·positiveOur 2026 Direction: AI and Classic Workflows in JetBrains IDEs
JetBrains outlines 2026 strategy to blend AI agents and traditional coding workflows within its IDEs without compromising either experience.
JetBrains Blog ↗
May 11·Opinion·positiveCode Maintainability: What is it and Why It’s Important
JetBrains Qodana blog explores code maintainability principles and static analysis tools to improve software quality and developer productivity.
JetBrains Blog ↗
May 11·Regulation·negativeTanStack Npm Packages Compromised Inside The Mini Shai Hulud Supply Chain Attack
Mini Shai-Hulud supply chain attack compromises 84 npm packages across TanStack, Mistral AI, and others via GitHub Actions exploit and OIDC token extraction.
Snyk (official) ↗
May 10·Launch·neutralWhy 157,000 developers are hedging against Anthropic with OpenCode
Anthropic doubled Claude Code rate limits and signed a major SpaceX deal for compute capacity, but faces competition from an open-source coding agent gaining developer adoption.
The New Stack ↗
May 10·Launch·positiveClaude can now follow users across Outlook, Word, Excel, and PowerPoint
Anthropic expands Claude into Microsoft 365 with new Outlook support and general availability for Word, Excel, and PowerPoint integrations.
The New Stack ↗
May 10·Opinion·neutralWhy Prometheus couldn’t see Cilium metrics at 2 a.m.
Grafana metrics visibility gap illustrates "integration tax" cost of coordinating multiple CNCF projects in production Kubernetes platforms.
The New Stack ↗
May 10·Launch·neutralAnthropic puts the “myth” in Mythos with its HackerOne bug bounty program
Anthropic launches public bug bounty program via HackerOne while restricting access to its more advanced Claude Mythos model to select security partners.
The New Stack ↗
May 9·Demo·neutralDatadog and T-Mobile leaders reveal the reality of deploying AI agents in production
Datadog extends observability tooling to validate and review AI-generated code before production deployment.
The New Stack ↗
May 9·Partnership·neutralHow Anthropic and Elon Musk cornered Sam Altman this week
Anthropic partners with Elon Musk's SpaceX to secure compute capacity at Colossus facility, boosting Claude Code capabilities amid intensifying AI infrastructure competition.
The New Stack ↗
May 8·Launch·positiveOpenAI Codex arrives in the browser with new Chrome extension
OpenAI launches Chrome extension for Codex to enable AI agents to operate within authenticated browser sessions without taking over the desktop.
The New Stack ↗
May 8·Launch·neutral“Several known limitations”: Developers react to Cursor’s promising but still-moving SDK
Cursor releases SDK for building AI agents using its own runtime and models, expanding beyond its IDE-only positioning.
The New Stack ↗
