A builder with one college coding class used Claude to ship Chipless, a free poker app tracking real cards and virtual chips. 19,072 hands dealt across 38 countries.
Claude Code v2.1.225 adds gateway spend-limit warnings and patches six bugs including 401 storms that broke headless and MCP OAuth sessions.
Starting August 14, Claude Code will use auto mode as its default permission setting for Pro, Max, and Team users. A separate classifier caught 89% of dangerous commands in testing.
Claude Code v2.1.224 adds self-hosted runners for Team/Enterprise, archive plugin installs, Bedrock region pinning, and JWT-aware credential masking in the sandbox layer.
Claude Code v2.1.223 patches five security issues including Bash permission bypass, Unicode command hiding, and workflow sandbox escapes, plus marketplace wildcards and /teleport hints.
Claude Code v2.1.221 patches a zsh permission bypass, adds VSCode Focus view for collapsing tool noise, and introduces credential masking for Linux sandboxes.
Claude Code v2.1.222 fixes worktree isolation letting subagents run destructive git commands, plus MCP usage attribution and proxy connectivity bugs.
Claude Code v2.1.207 ships auto mode without opt-in on Bedrock, Vertex, and Foundry, plus fixes for terminal lag, a security consent bypass, and a launcher overwrite bug.
Anthropic drops its policy of secretly degrading Claude Fable responses to frontier AI researchers after widespread backlash, promises visible safeguards instead.
Anthropic published a detailed engineering post on how it sandboxes Claude across Claude.ai, Claude Code, and Cowork — with gVisor, Seatbelt, Bubblewrap, and full VMs.
Simon Willison spent 5.5 hours stress-testing Anthropic's Claude Fable 5. It's a big model with big guardrails — and a new fallback mechanism when safety filters trigger.
Anthropic's Claude Opus 4.8 arrives with improved honesty, better uncertainty flagging, and benchmark leads over GPT-5.5 — and the company actually calls it a minor update.
Claude Opus 4.8 is live in Windsurf with unchanged standard pricing at $5/$25 per million tokens and a significantly cheaper Fast Mode option.
Anthropic is pulling Agent SDK and claude -p out of Pro and Max subscriptions on June 15, replacing the included usage with a fixed monthly API credit.
Anthropic is launching a standalone enterprise AI services company with Blackstone, Hellman & Friedman, and Goldman Sachs to deploy Claude across mid-market businesses.
Windsurf 2.0 adds Claude Opus 4.7, Anthropic's most capable model for coding and agentic tasks, with improved sustained reasoning for long sessions.
Anthropic launched Claude Design, a new tool for creating polished designs, prototypes, slides, and one-pagers through conversation with Claude.
Claude Code now supports Model Context Protocol servers, enabling direct integration with external tools and data sources from within the AI coding assistant.