The terminal is the new battleground. Google just killed Gemini CLI for Antigravity, xAI dropped Grok Build, and every major lab now wants to live in your shell.
The race isn't about who writes the best single line anymore. It's about who owns the agentic layer of your dev workflow. Plan mode, parallel subagents, CI pipelines. That's the new feature checklist.
Meanwhile, a new benchmark quietly reminded us agents still collapse on real databases. Good timing to ship the guardrails.
|
|
Together with
|
·
|
|
|
|
|
|
Today's Author
Lior Alexander. Founder of AlphaSignal and former ML Engineer. Previously built ML systems at Iguazio, Guesty, Enphase, Mila.
|
|
Top News
|
|
|
|
xAI opens Grok Build beta with image generation, video, and CLI automations
|
|
5,596 Likes
|
|
xAI just dropped Grok Build, a coding agent that lives in your terminal and actually takes action on your codebase. Not a chatbot. Not autocomplete. An agent that reads your project, makes a plan, edits files, and runs commands. The biggest problem with AI coding tools? They start executing and go off the rails before you notice. Grok Build solves this with Plan Mode: the agent shows you the full plan first, you approve each step, and nothing gets touched until you say so. Here is what you can do with it: - Run parallel subagents that split complex tasks and work simultaneously across your repo
- Use headless mode with a -p flag to plug it into CI pipelines and automated scripts
- Extend it with Skills, hooks, and MCP servers, and it already reads your existing CLAUDE.md with no changes needed
Install it with one command: curl -fsSL https://x.ai/cli/install.sh | bash. Available now for SuperGrok subscribers.
|
|
|
|
Presented by Lightfield
|
|
Outbound Agents Built on Your CRM
|
|
Most outbound tools run outside your stack - third-party lists, generic templates, zero context. Lightfield runs from the inside:
• Scores accounts against closed-won deal data • Generates sequences from real customer language • Maps warm intro paths across your network • Escalates only the replies that need you
You set the strategy. Agents handle list-building, sequencing, and follow-ups. 3,000+ startups on the platform.
|
|
|
|
Top Repo
|
|
|
|
Perplexity open-sources free security tool that protects developers from malicious AI coding tools
|
|
3,292 Likes
|
|
Hackers have been quietly poisoning free software packages that almost every app is built on. One group alone injected malicious code into over 160 packages, including a React tool with roughly 12 million weekly downloads. Install one, and attackers get a backdoor into everything you touch, including your AI tools. Perplexity just open-sourced their fix. It's called Bumblebee, the same tool they use internally to protect their own team. Here's what it scans: - Browser extensions across Chrome, Edge, Brave, Arc, and Firefox, plus editor plugins in VS Code and its forks
- MCP config files, the local settings that tell AI assistants which external services they can access, including emails, databases, and code repos
- Vulnerable packages across npm, PyPI, Go, and more, without executing package managers or touching your credentials
It uses a read-only scan, reading config files without running anything, so it can't accidentally trigger the malicious code it's looking for. Free, on GitHub, Apache 2.0 license.
|
|
|
|
Presented by Palabra
|
|
Real-time speech translation that sounds human.
|
|
Palabra.ai is a real-time AI speech translator built on a proprietary LLM: - Supports 60+ languages with voice cloning, speaker diarization and custom glossaries
- Works with Zoom, Meet and Teams in seconds, no setup required
- Built for global teams, events, broadcasters and developers.
|
|
|
|
Top Repo
|
|
|
|
ZOZO open-sources a physics solver that handles 180M clothing contacts without clipping
|
|
2,751 Likes
|
|
Ever seen a cloth simulation where a shirt clips through itself like a ghost? That's the problem ZOZO, Japan's largest fashion e-commerce company, just open-sourced a fix for. They released ppf-contact-solver, a physics simulation tool that handles how fabric, ropes, and soft objects touch and collide. The key thing: objects never clip through each other, and fabric never stretches beyond a strict limit you set, like 1% or 5%. The numbers are serious: - 180M+ contact points handled in a single scene, all running on the GPU
- Python API with JupyterLab built in, so you write simulation code in a browser notebook
- Docker image (~1GB), Windows .exe, and cloud-ready for AWS, GCP, RunPod
- Community Blender add-ons already available
You can rent a GPU for under $0.50/hour on vast.ai and run it remotely from your laptop. This opens up realistic cloth, rope, and soft-body simulation without needing expensive hardware or a game engine.
|
|
|
At Alpha Signal, our mission is to build a sharp, engaged community focused on AI, machine learning, and cutting-edge language models, helping over 200,000 developers stay informed and ahead. We're passionate about curating the best in AI, from top research and trending technical blogs to expert insights and tailored job opportunities. We keep you connected to the breakthroughs and discussions that matter, so you can stay in the loop without endless searching. We also work closely with partners who value the future of AI, including employers and advertisers who want to reach an audience as passionate about AI as we are.
Our partnerships are based on shared values of ethics, responsibility, and a commitment to building a better world through technology.Privacy is a priority at Alpha Signal. Our Privacy Policy clearly explains how we collect, store, and use your personal and non-personal information. By using our website, you accept these terms, which you can review on our website. This policy applies across all Alpha Signal pages, outlining your rights and how to contact us if you want to adjust the use of your information. We're based in the United States. By using our site, you agree to be governed by U.S. laws.
|
|
|
Looking to promote your company, product, service, or event to 250,000+ AI developers?
|
|
|
|
|