Dev Digest — March 10, 2026

🔥 HOT RELEASES

Anthropic — Claude Code Review Multi-agent code review system that dispatches parallel AI agents to review PRs, each examining from a different angle. A final agent aggregates findings, removes duplicates, and ranks by severity. Internal tests show it catches bugs 84% of the time on 1000+ line changes. $15-$25 per PR review. 🔗 X: https://x.com/rohanpaul_ai/status/2031403407182119318 🔗 TechCrunch: https://techcrunch.com/2026/03/09/anthropic-launches-code-review-tool-to-check-flood-of-ai-generated-code/

Anthropic — Claude Code Voice Mode Talk to your coding agent. Voice mode is rolling out gradually (~5% of users initially). Speak your instructions instead of typing — useful for pair-programming style workflows and quick architecture discussions. 🔗 TechCrunch: https://techcrunch.com/2026/03/03/claude-code-rolls-out-a-voice-mode-capability/

Cursor — Automations Always-on agentic coding. Automations trigger agents from events like new code pushes, Slack messages, or timers. Humans loop in only when judgment is needed. Includes built-in security auditing of diffs and agentic codeowners for PR risk classification. 🔗 TechCrunch: https://techcrunch.com/2026/03/05/cursor-is-rolling-out-a-new-system-for-agentic-coding/

OpenAI — Codex Native Windows App OpenAI's Codex AI coding tool finally gets a native Windows desktop app (previously Mac-only). 🔗 PCWorld: https://www.pcworld.com/article/3078509/openais-ai-coding-app-comes-to-windows-at-last.html

OpenAI Acquires Promptfoo OpenAI bought AI security startup Promptfoo (founded 2024) to bolster security testing for AI agents. Promptfoo's tools will integrate into OpenAI's Frontier platform. Signal: the "AI agent security stack" is becoming a real category. 🔗 TechCrunch: https://techcrunch.com/2026/03/09/openai-acquires-promptfoo-to-secure-its-ai-agents/ 💻 GitHub: https://github.com/promptfoo/promptfoo

Adalo — Ada AI that builds complete multi-screen apps from plain-language descriptions, then lets you visually direct changes. Idea to publishable product in seconds. 🔗 GlobeNewswire: https://www.globenewswire.com/news-release/2026/03/10/3252515/0/en/Adalo-Launches-Ada-AI-That-Builds-Complete-Multi-Screen-Apps-Then-Lets-You-Direct-It-Visually.html

🧪 INTERESTING REPOS

GitHub Security Lab — Taskflow Agent AI-powered vulnerability scanner from GitHub Security Lab. Finds auth bypasses, IDORs, token leaks automatically. Open-source framework for the kind of deep security auditing that usually requires expert humans. 🔗 Blog: https://github.blog/security/how-to-scan-for-vulnerabilities-with-github-security-labs-open-source-ai-powered-framework/ 💻 GitHub: https://github.com/github/security-lab

SecurityClaw Open-source SOC investigation tool — trending on X with 389 likes and 25K impressions. 🔗 X: https://x.com/Dinosn/status/2031206574736200090 💻 GitHub: https://github.com/SecurityClaw/SecurityClaw

InsForge 2.0 Tell your coding agent what you want and it spins up DB + Auth + deployment. Backend-as-a-service for agentic dev — fully open source. 🔗 X: https://x.com/CodeswithClara/status/2031411496060633362 💻 GitHub: https://github.com/InsForge/InsForge

🎥 WORTH WATCHING

"Claude Code Review, DeepSeek v4, Gemma 4, OpenClaw Update, Copilot Cowork & More!" Comprehensive AI news roundup covering all the major drops this week — Claude Code Review, DeepSeek v4 anticipation, Cursor Automations, and more. 🔗 YouTube: https://www.youtube.com/watch?v=td_lTMVV-ak

Washington Post — "We put Claude Cowork to the test" Interactive piece where WaPo readers submitted app ideas and Claude built them live. Good real-world demo of vibe coding capabilities and limits. 🔗 WaPo: https://www.washingtonpost.com/technology/interactive/2026/claude-code-cowork-build-apps/

💡 TECHNIQUES & IDEAS

AI "Convenience Loops" Reshaping Language Choice GitHub Octoverse data reveals a feedback loop: AI coding tools work best with statically typed languages → developers migrate to TypeScript → AI gets better at TypeScript → more migration. TypeScript surged 66% to become GitHub's #1 language. Types = guardrails for LLMs. 🔗 InfoQ: https://www.infoq.com/news/2026/03/ai-reshapes-language-choice/

Google Android Bench Google released a public leaderboard ranking AI models for Android app development. Gemini 3.1 Pro Preview leads at 72.4% success rate, followed by Claude and GPT models. Useful if you're picking models for mobile dev. 🔗 9to5Google: https://9to5google.com/2026/03/06/google-says-these-ai-models-are-best-at-coding-android-apps/

Amazon's AI-Linked Outage Amazon's site crashed on March 5 for ~6 hours — checkout, login, pricing all broken. 21K+ outage reports. Root cause: faulty code deployment where engineers used AI coding tools. Not the first time. Cautionary tale for shipping AI-generated code without proper review (makes Claude Code Review's timing perfect). 🔗 X: https://x.com/karankendre/status/2031434170598355394

🔮 EMERGING TRENDS

The $1B World Models Bet Yann LeCun's AMI Labs raised $1.03B at $3.5B valuation — Europe's largest seed round. Building "world models" that learn from physical reality instead of next-token prediction. Backed by Nvidia, Temasek, Bezos. Signal: investors betting on alternatives to LLMs for robotics/industrial AI. 🔗 Reuters: https://www.reuters.com/business/ex-meta-ai-chief-yann-lecuns-ami-raises-103-billion-alternative-ai-approach-2026-03-10/

OpenAI Building a GitHub Competitor Prompted by frequent GitHub outages disrupting their engineers, OpenAI is reportedly building an internal code-hosting platform. If it goes external, that's a direct competitor to Microsoft's GitHub — from Microsoft's biggest investment. 🔗 Tom's Hardware: https://www.tomshardware.com/tech-industry/openai-building-github-alternative-after-outages-disrupted-engineers

DeepSeek V4 — Still Imminent The highly anticipated multimodal model (text, image, video in one open model) with 1M token context keeps slipping. Was expected early March alongside China's "Two Sessions." Still hasn't dropped, but leaks suggest it's close. One to watch. 🔗 Reddit: https://www.reddit.com/r/DeepSeek/comments/1ridmnm/deepseek_v4_all_leaks_and_infos_for_the_release/

"Engineer" → "Builder" SF Standard reports on the cultural shift: with AI coding tools, engineers increasingly spend their days giving agents instructions in plain English and reviewing output. The industry is rebranding from "engineers" to "builders." 🔗 SF Standard: https://sfstandard.com/2026/03/05/engineer-2025-ai-land-everyone-s-builder-now/

Full digest: https://digest.99.cooking