Tag · 36 stories
Every story tagged Claude Code on AI Chat Daily.
Claude Code is Anthropic's CLI-native coding agent and the reference product for AI-assisted software engineering in 2026. We follow every Claude Code release, Agent SDK updates, and how developers are using it in practice.
Anthropic confirms infostealer malware is hijacking Claude Code OAuth tokens; one Max user watched usage climb from 45% to 55% while doing no work.
Meta's new Muse Spark model charges 10 cents per million input tokens if you let it train on your prompts and outputs.
A single April tweet from Andrej Karpathy sent thousands of researchers building Markdown wikis maintained by Claude and GPT-5.5.
The system packages agent orchestration, evals, and CI-style tooling into a single layer aimed at teams that can't build Stripe's setup from scratch.
Starting August 14, Pro, Max, and Team accounts get an agent that stops asking permission at every step.
Augment says its context engine runs Terminal-Bench 33% more efficiently than Claude Code using the same underlying model.
Signed by Anthropic, Google, and OpenAI executives alongside Nobel laureates, the statement says AI could transform the economy faster than the Industrial Revolution.
Beijing's cybersecurity platform tells users to uninstall or upgrade Claude Code versions 2.1.91 through 2.1.196, released April 2 to June 29.
Raycast's app-making app now available to everyone, with a Pro tier offering 200 credits monthly for building personal desktop software.
The Chinese cloud giant is cutting off internal access to Claude Code as US export controls tighten around frontier AI tools.
The launch follows October's Cursor 2.0 agent overhaul and mirrors mobile coding moves from Anthropic and OpenAI.
The design platform pulls coding, animation, and Weavy workflows into one collaborative surface as it pushes deeper into AI tooling.
The Claude Code creator says swarms of agents prompting agents are as big a leap as the move from hand-written code to agentic AI.
Teams of up to 8 agents from OpenAI, Anthropic and Moonshot taught robots to insert GPUs and cut zip ties without human input.
Wharton's Ethan Mollick says the new Mythos-class model executed multi-page specs autonomously for up to a dozen hours.
Uber burned its full-year AI coding budget by April. One company racked up a $500M Claude bill. The FinOps reckoning has arrived.
A subreddit of 140,000 soccer fans is using Claude Code to scrape FIFA's marketplace, expose scalper markups, and route resales through WhatsApp.
The ridesharing company now meters Claude Code and Cursor usage on an internal dashboard after blowing through its AI budget.
The new Opus holds Opus 4.7 pricing at $5/$25 per million tokens, while fast mode drops to a third of its prior cost at 2.5x speed.
Anthropic's Claude Code creator hasn't written code in six months and predicts 100x more builders.
The company exhausted its annual AI budget four months into 2026 with no clear link between token usage and feature delivery.
Almost half the room at Code with Claude raised their hands when asked if they'd shipped Claude-written code unreviewed.
The design platform's new agent generates, edits, and automates designs in real time alongside human teammates.
Cat Wu says the Claude Code team ships on week-long cycles and strips out scaffolding as models improve, betting form factor will collapse.
After six months of internal Claude Code use, Microsoft's Experiences + Devices group is winding down access by June 30 — the last day of its fiscal year.
The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.
The briefing read inside teams at
