Claude Opus 4.7 Now Available
Claude Opus 4.7 is now available in Claude Code with a new xhigh effort level, alongside a native binary for faster startup and a parallel multi-agent code review command. The release arrives as Anthropic fields investor interest at $800 billion, negotiates model access with the White House, and shifts enterprise customers to usage-based pricing.
White House and Anthropic Hold 'Productive' Meeting, Aiming for a Compromise — New York Times
Highlight
Claude Opus 4.7 Now Available With New xhigh Effort Level
You can now use Opus 4.7 in Claude Code, including a new xhigh effort level that sits between high and max. Run /effort without arguments to get an interactive slider for picking the right speed-intelligence tradeoff. If you're a Max subscriber, auto mode now works with Opus 4.7 and no longer requires the --enable-auto-mode flag. Anthropic also published a companion blog post on best practices for Opus 4.7 covering recalibrated effort levels and adaptive thinking.
Worth Knowing
CLI Now Ships as a Native Binary
Claude Code now spawns a platform-specific native binary instead of running bundled JavaScript through Node. You should see faster startup times as a result. The change is transparent -- your existing commands and flags work the same way.
Fullscreen TUI Mode and Readline-Style Keybindings
You can now run /tui fullscreen mid-conversation to switch to a flicker-free fullscreen renderer. The update also brings readline-style keybindings: Ctrl+A and Ctrl+E jump to the start and end of the current line, Ctrl+Backspace deletes the previous word, and Shift+Up/Down scrolls the viewport when extending a selection past the visible edge.
You Can Now Run Parallel Code Reviews With /ultrareview
Run /ultrareview to launch a multi-agent code review that analyzes your current branch in parallel, or pass a PR number to review a specific GitHub pull request. The review runs in the cloud, so it doesn't block your local session.
New /less-permission-prompts Command Reduces Approval Fatigue
If you're tired of approving the same read-only commands, run /less-permission-prompts to scan your recent transcripts and generate a proposed allowlist for .claude/settings.json. Separately, read-only commands with glob patterns (like ls *.ts) and commands prefixed with cd <project-dir> && no longer trigger prompts at all.
Session Recaps Now Summarize Where You Left Off
When you return to a session after being away, Claude Code now shows a recap of what happened. You can also invoke it manually with /recap. The feature now works even if you have telemetry disabled -- opt out via /config or CLAUDE_CODE_ENABLE_AWAY_SUMMARY=0.
Under the Hood
Bash permission rules hardened against wrapper commands, find -exec, and sandbox bypass
Your Bash deny rules now catch commands wrapped in env, sudo, watch, and similar exec wrappers. Bash(find:*) allow rules no longer auto-approve find -exec or -delete, macOS symlinked system paths are treated as dangerous rm targets, and a sandbox bypass that allowed dangerouslyDisableSandbox to skip permission prompts has been fixed.
Terminal rendering fixes for iTerm2, tmux, Ghostty, Kitty, and others
If you use iTerm2 with tmux, Ghostty, Kitty, Alacritty, or WezTerm over SSH/mosh, you should see fewer display glitches -- tearing, garbled output, and washed-out colors have all been addressed. Memory usage also dropped thanks to on-demand loading of language grammars for syntax highlighting.
From Anthropic
In the News
Between the Opus 4.7 launch, the native binary migration, and the Bash security hardening, this is a week where updating early pays off. If you haven't tried the new /effort slider or /ultrareview, both are worth a look.