Skip to main content
Back

Anthropic Claude Security Beta: AI scans find bugs missed for years

Anthropic launches Claude Security, Cursor releases agent SDK, Ant Group open-sources 1T-param model, agent optimizer boosts AppWorld score to 89.5

May 1, 2026 6 min read

Hey,

Today's theme is simple: AI is done waiting.

Claude Security ships to prod, catching bugs that sat undetected for years. KAME, a new voice architecture, starts talking before it finishes thinking. Cursor opens its agent SDK. The pattern? Every layer of the stack is collapsing the gap between "thinking" and "doing."

Speed without sloppiness. That's the new bar.
Together with · partner_image
Author photo
Today's Author
Lior Alexander. Founder of AlphaSignal and former ML Engineer. Previously built ML systems at Iguazio, Guesty, Enphase, Mila.
Summary

Read time: 6 min 30 sec

Top News

▸ Anthropic launches Claude Security to scan codebases and fix vulnerabilities without custom tooling

Tiger Data

▸ Your Agent Needs More Than 2 Projects

Top Repo

▸ Cursor releases open-source SDK to build and deploy coding agents anywhere

Braintrust

▸ Ship quality AI like Ramp, Notion, and Lovable

Top Repo

▸ Nous Research's Hermes agent can now run ComfyUI workflows on demand

Signals

1. Open-source agent optimizer boosts Claude's AppWorld score from 73.7 to 89.5
2. Chip Huyen releases free resources alongside her 2025 AI Engineering book
3. New tandem voice model speaks while thinking, cutting latency without losing accuracy
4. IBM Granite 4.1 30B model gets quantized GGUF builds for local use
5. Ant Group open-sources a 1 trillion-parameter model that skips token-wasting reasoning
Top News
Anthropic launches Claude Security to scan codebases and fix vulnerabilities without custom tooling
16,811 Likes
Anthropic launches Claude Security to scan codebas

Claude Security just hit public beta, and it's the easiest way to add AI-powered vulnerability scanning to your codebase. No API integration or custom agent build required: if your org already uses Claude, you can start scanning today.

Here's what makes it different from old-school scanners. Traditional tools use rule-based pattern matching, which catches known vulnerability patterns but misses complex issues. Claude Security focuses on memory corruption, injection flaws, auth bypasses, and complex logic errors that pattern-matching tools typically miss.

The noise problem is handled too. Claude assesses code like a human researcher, verifies findings before they reach you, and gives each result a confidence rating with severity and likely impact.

What you get out of the box:

  • Scheduled scans on a full repo or a specific directory, no manual kickoffs
  • A ready-to-review patch you can open directly in Claude Code
  • Push findings to Slack, Jira, or any ticketing system via webhooks, or export as CSV or Markdown

Hundreds of organizations used it to fix issues in production code that existing tools had missed for years. Available now on Enterprise.

Read more
forward →
Presented by Tiger Data
Your Agent Needs More Than 2 Projects

Most Postgres providers give you a handful of databases and call it a free tier.

Your agent burns through that before lunch.

Ghost is built for how agents actually use databases:

  • disposable
  • isolated
  • burst-oriented

Fork production in 58ms. Spin up a fresh instance per task. Delete without cleanup. No project limits, no shared state, no naming conventions that nobody follows. Same SQL. Same tooling. Different lifecycle.

100 hours/month free. 1TB storage. No credit card.

Start building
partner with us →
Top Repo
Cursor releases open-source SDK to build and deploy coding agents anywhere
8,309 Likes
Cursor releases open-source SDK to build and deplo

Cursor just opened up its AI coding agent to the outside world. Until now, those agents only lived inside the Cursor editor. Now you can run them anywhere.

Install it with npm install @cursor/sdk and you get the exact same agent engine that powers Cursor itself. Not a watered-down version. The real thing.

What can you actually do with it?

  • Hook it into your CI/CD pipeline so agents auto-fix failing tests and open a pull request, no human needed
  • Embed a coding agent inside your own product or internal tool
  • Spin up background agents that take a bug ticket all the way to a merge-ready PR
  • Run agents in Cursor's cloud, where each gets its own sandboxed VM with your repo already cloned, and keeps running even if you go offline

Billing is token-based, so you pay for what you use. Starter projects are on GitHub at cursor/cookbook to get you moving fast.

Read more
forward →
Presented by Braintrust
Ship quality AI like Ramp, Notion, and Lovable

Ramp, Notion, and OpenAI use Braintrust to ship quality AI and catch regressions before users do.

In Braintrust's 1-hour course, you will learn how to build a chatbot from scratch, write scorers, compare prompts, and turn production traces into tests.

Start the course
partner with us →
Top Repo
Nous Research's Hermes agent can now run ComfyUI workflows on demand
3,233 Likes
Nous Research's Hermes agent can now run ComfyUI w

Hermes Agent just got a big upgrade: it can now control ComfyUI end-to-end. ComfyUI is an open-source tool for generating images, video, and audio using visual, node-based pipelines. Normally, running it means manually installing it, setting up models, and wiring workflows yourself. That friction is now gone.

Hermes is an open-source AI agent by Nous Research that lives on your server, remembers what it learns, and gets more capable over time. Think of it as a personal assistant that runs in your terminal or on Telegram.

With this update, you can ask Hermes in plain English to generate media, and it handles the rest. Here is what it can do:

  • Install and launch ComfyUI automatically
  • Manage custom nodes and models on demand
  • Run full ComfyUI workflows with parameter injection via REST and WebSocket

ComfyUI support is now built-in by default, not an optional add-on. To get started: git clone the repo and run ./setup-hermes.sh.

Read more
forward →
Signals
1
Open-source agent optimizer boosts Claude's AppWorld score from 73.7 to 89.5
1,246 Likes
2
Superwhisper launches Claude Code integration. Run agents in parallel, get notified when they need you, dictate a reply.
Presented by SuperWhisper
3
Chip Huyen releases free resources alongside her 2025 AI Engineering book
15,428 Stars
4
New tandem voice model speaks while thinking, cutting latency without losing accuracy
542 Likes
5
IBM Granite 4.1 30B model gets quantized GGUF builds for local use
1,305 Downloads
6
Ant Group open-sources a 1 trillion-parameter model that skips token-wasting reasoning
982 Likes
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?
Work with us