AI News
Signal over noise. Daily.
一線 AI Lab
19 articlesIntroducing Claude Tag
Claude Tag is a new way for teams to work with Claude.
Scale to many tools with tool search
Scale your agent to thousands of tools by discovering and loading only what's needed, on demand. Tool search enables your agent to work with hundreds or thousands of tools by dynamically discovering and loading them on demand. Instead of loading all tool definitions into the context window upfront, the agent searches your tool catalog and loads only the tools it needs. This approach solves two challenges as tool libraries scale: * **Context efficiency:** Tool definitions can consume large portions of the context window (50 tools can use 10-20K tokens), leaving less room for actual work. * **Tool selection accuracy:** Tool selection accuracy degrades with more than 30-50 tools loaded at once. Tool search is enabled by default. When tool search is active, tool definitions are withheld from…
Agent SDK reference - TypeScript
Complete API reference for the TypeScript Agent SDK, including all functions, types, and interfaces. npm install @anthropic-ai/claude-agent-sdk The SDK bundles a native Claude Code binary for your platform as an optional dependency such as `@anthropic-ai/claude-agent-sdk-darwin-arm64`. You don't need to install Claude Code separately. If your package manager skips optional dependencies, the SDK throws `Native CLI binary for not found`; set `pathToClaudeCodeExecutable` to a separately installed `claude` binary instead. When you compile your application into a single-file executable with `bun build --compile`, the SDK cannot resolve the bundled CLI binary at runtime. `require.resolve` does not work inside the compiled executable's `$bunfs` virtual filesystem, so the SDK throws `Native CLI…
Agent SDK reference - Python
Complete API reference for the Python Agent SDK, including all functions, types, and classes. Install the package into a virtual environment. On recent Debian, Ubuntu, and Homebrew Python installs, running `pip install` against system Python fails with `error: externally-managed-environment`. python3 -m venv .venv source .venv/bin/activate pip install claude-agent-sdk For uv, Windows PowerShell, and API key setup, see Get started in the Agent SDK overview. The Python SDK provides two ways to interact with Claude Code: | Feature | `query()` | `ClaudeSDKClient` | | :------------------ | :--------------------------------------------- | :--------------------------------- | | **Session** | Creates a new session by default | Reuses same session |
Configure permissions
Control how your agent uses tools with permission modes, hooks, and declarative allow/deny rules. The Claude Agent SDK provides permission controls to manage how Claude uses tools. Use permission modes and rules to define what's allowed automatically, and the `canUseTool` callback to handle everything else at runtime. This page covers permission modes and rules. To build interactive approval flows where users approve or deny tool requests at runtime, see Handle approvals and user input. When Claude requests a tool, the SDK checks permissions in this order: Run hooks first. A hook can deny the call outright or pass it on. A hook that returns `allow` does not skip the deny and ask rules below; those are evaluated regardless of the hook result.
Track cost and usage
Learn how to track token usage, estimate costs, and configure prompt caching with the Claude Agent SDK. The Claude Agent SDK provides detailed token usage information for each interaction with Claude. This guide explains how to properly track usage and understand cost reporting, especially when dealing with parallel tool uses and multi-step conversations. For complete API documentation, see the TypeScript SDK reference and Python SDK reference. The `total_cost_usd` and `costUSD` fields are client-side estimates, not authoritative billing data. The SDK computes them locally from a price table bundled at build time, so they can drift from what you are actually billed when: * pricing changes * the installed SDK version does not recognize a model * billing rules apply that the client cannot…
How the agent loop works
Understand the message lifecycle, tool execution, context window, and architecture that power your SDK agents. The Agent SDK lets you embed Claude Code's autonomous agent loop in your own applications. The SDK is a standalone package that gives you programmatic control over tools, permissions, cost limits, and output. You don't need the Claude Code CLI installed to use it. When you start an agent, the SDK runs the same execution loop that powers Claude Code: Claude evaluates your prompt, calls tools to take action, receives the results, and repeats until the task is complete. This page explains what happens inside that loop so you can build, debug, and optimize your agents effectively. Every agent session follows the same cycle:
Handle approvals and user input
Surface Claude's approval requests and clarifying questions to users, then return their decisions to the SDK. While working on a task, Claude sometimes needs to check in with users. It might need permission before deleting files, or need to ask which database to use for a new project. Your application needs to surface these requests to users so Claude can continue with their input. Claude requests user input in two situations: when it needs **permission to use a tool** (like deleting files or running commands), and when it has **clarifying questions** (via the `AskUserQuestion` tool). Both trigger your `canUseTool` callback, which pauses execution until you return a response. This is different from normal conversation turns where Claude finishes and waits for your next message. For…
Run Claude Code programmatically
Use the Agent SDK to run Claude Code programmatically from the CLI, Python, or TypeScript. The Agent SDK gives you the same tools, agent loop, and context management that power Claude Code. It's available as a CLI for scripts and CI/CD, or as Python and TypeScript packages for full programmatic control. To run Claude Code in non-interactive mode, pass `-p` with your prompt and any CLI options: claude -p "Find and fix the bug in auth.py" --allowedTools "Read,Edit,Bash" This page covers using the Agent SDK via the CLI (`claude -p`). For the Python and TypeScript SDK packages with structured outputs, tool approval callbacks, and native message objects, see the full Agent SDK documentation. Add the `-p` (or `--print`) flag to any `claude` command to run it non-interactively. All CLI options…
Connect to external tools with MCP
Configure MCP servers to extend your agent with external tools. Covers transport types, tool search for large tool sets, authentication, and error handling. The Model Context Protocol (MCP) is an open standard for connecting AI agents to external tools and data sources. With MCP, your agent can query databases, integrate with APIs like Slack and GitHub, and connect to other services without writing custom tool implementations. MCP servers can run as local processes, connect over HTTP, or execute directly within your SDK application. This page covers MCP configuration for the Agent SDK. To add MCP servers to the Claude Code CLI so they load in every project, see MCP installation scopes. This example connects to the Claude Code documentation MCP server using HTTP transport and uses…
Quickstart
Get started with the Python or TypeScript Agent SDK to build AI agents that work autonomously Use the Agent SDK to build an AI agent that reads your code, finds bugs, and fixes them, all without manual intervention. **What you'll do:** 1. Set up a project with the Agent SDK 2. Create a file with some buggy code 3. Run an agent that finds and fixes the bugs automatically * **Node.js 18+** or **Python 3.10+** * An **Anthropic account** (sign up here) Create a new directory for this quickstart: mkdir my-agent cd my-agent For your own projects, you can run the SDK from any folder; it will have access to files in that directory and its subdirectories by default. Install the Agent SDK package for your language:
Finance Agents
We're releasing ten new Cowork and Claude Code plugins, integrations with the Microsoft 365 suite, new connectors, and an MCP app for financial services and insurance organizations.
Use Claude Code on the web
Configure cloud environments, setup scripts, network access, and Docker in Anthropic's sandbox. Move sessions between web and terminal with `--cloud` and `--teleport`. Claude Code on the web is in research preview for Pro, Max, and Team users, and for Enterprise users with premium seats or Chat + Claude Code seats. Claude Code on the web runs tasks on Anthropic-managed cloud infrastructure at claude.ai/code. Sessions persist even if you close your browser, and you can monitor them from the Claude mobile app. New to Claude Code on the web? Start with Get started to connect your GitHub account and submit your first task. This page covers: * GitHub authentication options: two ways to connect GitHub * The cloud environment: what config carries over, what tools are installed, and how to…
What building Shippy taught us about building agents
A Blog post by Ai2 on Hugging Face
Model Routing Is Simple. Until It Isn’t.
A Blog post by IBM Research on Hugging Face
Claude For Teachers
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
The US is advancing AI safety through state and federal action
OpenAI outlines a “reverse federalism” approach to AI governance, where state laws help build a national framework for safe, democratic AI.
GPT-Red: Unlocking Self-Improvement for Robustness
Explore GPT-Red, OpenAI’s automated red teaming system that uses self-play to improve AI safety, alignment, and prompt injection robustness.
Canadian AI Research
Anthropic is committing $10M to Canadian research institutions to fund the next generation of AI research.
媒體
32 articlesMicrosoft is reportedly training salespeople to talk down OpenAI and Anthropic
Microsoft is looking to sell its in-house AI models as more efficient and cost-effective than its competitors' models.
xAI sues a man for using Grok to generate CSAM ‘deepfakes’
The Elon Musk-owned xAI is suing a South Carolina man who allegedly used the company's Grok AI chatbot to generate child sexual abuse material (CSAM). In a lawsuit reported earlier by Reuters , xAI claims Terry Wayne Harwood "knowingly and intentionally used Grok to circumvent safeguards, alter nonconsensual images, and generate and distribute CSAM," breaching the company's policies. Harwood was arrested in February for allegedly possessing and distributing CSAM and is facing eight felony charges. The lawsuit claims "at least some" of the images related to Harwood's criminal charges "were generated or altered" with Grok, and that Harwood byp … Read the full story at The Verge.
SpaceX Briefly Falls Below I.P.O. Price of $135
After soaring in its stock market debut, the shares of Elon Musk’s rocket and A.I. company have been rocky.
AI slop movies are the new direct-to-video cash grabs
This weekend, cinephiles across the world will march to their local theaters to feast their eyes on Christopher Nolan's new adaptation of The Odyssey . It's on track to rake in anywhere between $80-$100 million in just a few days. People are clearly excited to see how Nolan uses cutting-edge filmmaking tech to make the Homeric classic feel fresh. But another director is trying to capitalize on the buzz around Nolan's project to drum up interest in an Odysseus-focused movie of his own. On Tuesday, film studio Fountain 0 announced that it is working on an AI-generated reimagining of The Odyssey titled Odysseus: The Fall , which will be availabl … Read the full story at The Verge.
Amid hardware legal battle, OpenAI releases a $230 keyboard for Codex
OpenAI, which is in the middle of a legal battle with Apple over hardware trade theft allegations, just released a light-up keyboard designed to be paired with its agentic coding app.
AI Isn’t Smarter Than a Baby—Yet
Babies are tremendous learning machines, and key advances for AI may soon be found in the architecture of their little brains.
China’s Chip Champion to Raise Billions in Race for A.I. Control
CXMT, a linchpin in China’s drive to develop homegrown artificial intelligence technology, is aiming to raise nearly $10 billion in a blockbuster public offering in Shanghai.
SpaceX falls to $135 IPO price ahead of Starship launch
The stock has steadily fallen from the euphoric post-IPO high, showing that markets may be sobering up to the promises CEO Elon Musk made before and after SpaceX went public.
Thinking Machines Lab Drops Its First Model
Inkling, a 975-billion-parameter open source model, was trained to understand video and audio. It could help Thinking Machines establish itself among competitors like Anthropic and OpenAI.
Thinking Machines amps up its bet against one-size-fits-all AI with its first open model, Inkling
It's the company's first public proof point after a year and a half spent building AI infrastructure largely out of public view.
China’s Economy Grows at Slowest Pace in Years
Economic growth of 4.3 percent in the second quarter, versus the same period last year, reflected a broad slump outside of the country’s export-oriented manufacturing might.
Suno snatched millions of songs from YouTube, Genius, and Deezer
So when is fair use actually just stealing? | Cath Virginia / The Verge | Photo from Getty Images Suno data obtained in a hacking incident has exposed that the AI music generator was trained by scraping millions of songs and lyrics from online audio platforms, including YouTube Music, Deezer, and Genius, 404 Media reports. Given that Suno has avoided revealing what's in its training datasets and how they were acquired, this a rare glimpse into what Suno has actually been taking from online platforms. That's relevant because Suno has been the subject of several lawsuits that allege it used copyrighted materials to train its AI models. In a notable case filed by the Recording Industry Association of America (RIAA), Suno openly admitted th … Read the full story at The Verge.
Meet GPT-Red: an LLM super-hacker OpenAI built to make its models safer
OpenAI has built an LLM super-hacker called GPT-Red that it uses as a sparring partner to help its other models boost their defenses against cyberattacks. Last week the company released the latest version of its flagship LLM, GPT-5.6. OpenAI says that training it against GPT-Red made the model its most robust release yet. GPT-Red automates a type of safety evaluation for software systems known as red-teaming, which is typically done by a team of human testers. The aim is to find as many different ways to break or hijack a system as possible. The weak spots can then be patched before the final version of the software is released. As LLMs become more complex and get used in a wider variety of tasks—especially in the form of agents, which can interact with computer files, websites, and…
Hack suggests AI music generator Suno scraped YouTube for training data
The hacker used an employee's credentials to access source code, which revealed how Suno scraped decades of audio.
Whatnot acquires Shaped to power real-time live shopping recommendations
Livestream shopping platform Whatnot has acquired AI startup Shaped, a machine learning company focused on real-time recommendations and search. The deal will bolster Whatnot’s personalization and discovery features as it expands into new product categories.
Microsoft patches record number of security vulnerabilities, citing its use of AI
Microsoft's monthly release of security fixes, dubbed Patch Tuesday, resolved a record 570 security vulnerabilities across the company's product line, thanks to discoveries with AI.
OpenAI's first branded hardware is... a light-up keyboard?
As rumors continue to swirl about OpenAI's work on a personalized smart speaker and other hardware, the company is today rolling out its first branded device. The $230 Codex Micro is a specialized, RGB-lit mini-keyboard designed to let users monitor and quickly interact with multiple Codex agents with a glance and a few clicks. The device is described as a "limited-run collaboration" with Work Louder , which already sells a very similar-looking Creator Micro line of customizable square keyboards targeted at creative professionals. The Codex Micro differentiates itself from those mainly through six frosted keys in the top two rows, which offer color-coded live feedback on up to six Codex threads, even when they are not in focus on-screen. Open the OpenAI box for AI assistance. Credit:…
OpenAI finally launches hardware… for Codex
OpenAI is finally releasing some hardware. No, it isn't the mysterious AI-powered device the company is developing with former Apple designer Jony Ive , a project already tangled up in a messy lawsuit . Instead, it's a product designed to be used with its coding platform, Codex. The device, a square-shaped block of buttons called Codex Micro, is a collaboration between the AI company and keyboard maker Work Louder. OpenAI said it is a limited-run collaboration that will give users more ways to monitor and manage their agents. The pad closely resembles Work Louder's Creator Micro 2 , and marketing images show what appears to be an identical … Read the full story at The Verge.
Apple Intelligence approved for launch in China with Alibaba’s Qwen AI
The deal, which was rumored to be in the works last year, marks an important step for Apple's AI ambitions in a key market.
Inside Ode with Anthropic, the startup betting AI services are the future of enterprise
Can a handful of engineers really do the work of an army of consultants? That’s the bet behind Ode with Anthropic — the joint venture dedicated to embedding forward-deployed engineers in enterprise firms, backed by Anthropic, Blackstone, Hellman & Friedman, Goldman Sachs and others. On this episode of TechCrunch’s Equity podcast, Rebecca Bellan sits down with Ode’s leaders Chris Taylor and Eddie Siegel, who founded Fractional AI, […]
Australia to Impose Energy and Water Guardrails on Data Centers Amid A.I. Boom
The country will also seek to protect the rights of creators of work used to train artificial intelligence models, as it aims to impose parameters on the growing industry.
Is This the Fastest Opinion Shift in American Politics?
David Wallace-Wells and Robinson Meyer discuss whether the gathering coalition of Americans against data centers could be the next Occupy Wall Street movement or Tea Party.
Anthropic, Blackstone bet the next trillion-dollar AI business is implementation, not just models
Anthropic-backed Ode launches as AI labs bet that embedding forward-deployed engineers inside enterprises is the key to accelerating enterprise AI adoption.
Reelful’s AI turns your camera roll into short-form videos for social media
The app is designed for people who want to create social content, but find traditional video editing tools too complex or time-consuming.
Rime picks up $24M Series A to help enterprises field customer calls
Rime is handling over 100 million calls each month across multiple companies.
IBM’s Stock Crash Reignites Fears of A.I. Disruption in Software Sector
The company tumbled sharply after previewing quarterly results that were disappointing. It’s not good news for software and tech consulting companies.
Indian AI coding startup Emergent becomes a unicorn with $130M Series C
The startup has reached a $120 million annualized revenue run rate and more than 200,000 paying customers.
Vint Cerf is working on a plan to unleash AI agents on the open internet
The guy behind TCP/IP is working on a standard for identifying AI agents in the wild.
An Inventor of Apple’s FaceID Wants to Analyze Your Brain’s Health With AI
Gidi Littwin’s new AI startup, Hemispheric, makes diagnostic brain scans for conditions like depression, PTSD, and Parkinson’s. He wants the technology to be as cheap and easy as for a blood test.
My Ebike Delivery Went Missing. When I Tried to Recover It, I Ended Up in Chatbot Hell
Companies’ increasing reliance on AI chatbots isn’t making the customer service experience smarter. It’s just making it more infuriating.
OpenAI Staffers Are Funding a Rival Super PAC to Take on Their Boss
OpenAI employees have donated more than $215,000 to a political effort opposing Leading the Future, a group backed by the company’s president, Greg Brockman.
Data Centers to Add Billions in Power Costs in 13 States
A power auction conducted by a giant grid operator is expected to add $6.3 billion in additional charges to consumers and businesses because of electricity needs of data centers.