v2.1.113

2025-02-22 · Source: Claude Code Changelog · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Cybersecurity & Data Privacy · Depth: Intermediate, long

Summary

Claude Code versions 2.1.105 through 2.1.114 introduce numerous fixes, security enhancements, and new features, significantly improving the user experience and agent capabilities. Key updates include the availability of Claude Opus 4.7 xhigh with an interactive `/effort` slider, and the introduction of "Auto mode" for Max subscribers. New tools like `/ultrareview` enable comprehensive, parallelized code review, and `/less-permission-prompts` helps generate Bash and MCP tool allowlists. Security improvements focus on stricter `Bash` deny rules, safer handling of macOS `/private` paths, and preventing command injection. Performance enhancements include faster `/ultrareview` launches, improved API stream handling, and reduced memory footprint for file operations. Various bug fixes address issues ranging from UI rendering and input handling to plugin management and remote control client functionality.

Key takeaway

For NLP Engineers and CTOs evaluating or deploying Claude Code, these updates significantly enhance security and operational efficiency. The introduction of Opus 4.7 xhigh and "Auto mode" offers more powerful and flexible model interaction, while `/ultrareview` and improved permission management streamline development workflows. You should review the new security features and consider adopting `/less-permission-prompts` to harden your agent environments and simplify permission configurations.

Key insights

Recent Claude Code updates enhance security, performance, and agent capabilities through new features and bug fixes.

Principles

• Prioritize security in tool execution.
• Optimize for user experience and efficiency.
• Provide clear feedback and error handling.

Method

The `/ultrareview` command leverages parallel multi-agent analysis and critique for comprehensive code review, while `/less-permission-prompts` scans transcripts to propose prioritized allowlists for `.claude/settings.json`.

In practice

• Use `/effort` to tune Opus 4.7 speed vs. intelligence.
• Employ `/ultrareview` for comprehensive code reviews.
• Configure `sandbox.network.deniedDomains` to block specific domains.

Topics

• Claude Opus 4.7
• AI Assistant Features
• Security Enhancements
• User Interface
• Developer Workflow

Best for: NLP Engineer, CTO, VP of Engineering/Data, AI Engineer, Machine Learning Engineer, Software Engineer

Related on AIssential

• v2.1.112 — Recent Claude Code CLI updates enhance security, performance, and user experience through new features and bug fixes.
• v2.1.101 — Recent Claude Code updates enhance security, user experience, and model capabilities, including Opus 4.7 xhigh and advanced code review.
• Claude Security is now in public beta — Claude Security uses Opus 4.7 to find and fix code vulnerabilities, accelerating security workflows for enterprises.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Claude Code Changelog.