Claude Code /code-review: Renamed from /simplify with Optional Effort Levels
Claude CodeClaude Code v2.1.146 renames the /simplify command to /code-review, giving it a name that accurately reflects its purpose — reviewing changed code for reuse, quality, and efficiency. The command now accepts an optional effort level argument (e.g., /code-review high), allowing developers to dial in how thorough the review should be. The same release also fixes a significant behavioral regression in auto mode where AskUserQuestion calls were being suppressed even when a skill explicitly depended on them, and improves diff rendering performance for large file edits.
Sources & Mentions
3 external resources covering this update
New Features
/code-review: A More Honest Name for Code Analysis
Claude Code v2.1.146 retires the /simplify slash command in favor of /code-review, a rename that better describes what the command actually does. Rather than just simplifying code, it performs a structured review of changed code — identifying over-engineering, catching quality issues, and surfacing opportunities to improve reuse and efficiency.
The rename also comes with a meaningful capability addition: the command now accepts an optional effort level argument. Running /code-review high instructs Claude Code to apply deeper analysis, while a bare /code-review runs a standard pass. This mirrors the effort-level system used elsewhere in Claude Code and gives developers explicit control over the depth-versus-speed tradeoff.
For teams already using the Code Review plugin, this built-in command offers a lightweight alternative that does not require plugin installation — it is available to all Claude Code users immediately after updating.
Auto Mode No Longer Blocks AskUserQuestion for Skills That Need It
A behavioral fix in this release addresses a frustrating regression in auto mode: Claude Code was suppressing AskUserQuestion calls even when the active user or skill explicitly depended on them to gather input. In practice, this meant agents and skills that used AskUserQuestion as a core interaction pattern would silently fail to collect required information when running under auto mode.
The fix restores the expected behavior: when a user or skill explicitly relies on AskUserQuestion, auto mode no longer overrides it. This is particularly impactful for teams building custom skills or running agentic workflows where human confirmation is a deliberate checkpoint.
Performance Improvements
Faster Diff Rendering for Large File Edits
Claude Code v2.1.146 includes improved diff rendering performance specifically for large file edits. For developers working in large codebases or making sweeping refactors that touch thousands of lines, the diff display now renders significantly faster. This is a quality-of-life improvement with no configuration required — it applies automatically to all file edits.
Key Reliability Improvements
The auto-updater also receives a reliability upgrade in this release: version checks and downloads now retry transient network failures instead of failing immediately. For users on intermittent or restricted networks (common in corporate or VPN environments), this means fewer failed update attempts and less manual intervention needed to stay on the latest version.