Anthropic has released 10 major updates to Claude Code over the past month that dramatically improve workflow efficiency. Key features include session resumption (claude --res), context visualization (/context), usage stats (/stats), session naming (/rename), easy rewind (double Escape), Ultra Think mode for complex tasks, custom memories via hashtag, YOLO mode for experienced users, a new plugin store (/plugins), and prompt stashing (Ctrl+S).
claude --res to bring back all context from previous conversations - no more losing work when you restart/context to visualize what's filling your context window and use /clear to free up space/stats shows tokens used, streaks, sessions, active days, and model breakdown/rename to label sessions for easy identification when running multiple sessionsclaude --dangerously-skip-permissions for autonomous operation (experienced users only)/plugins for one-click installation of skills, MCPs, and integrationsThe claude --res command allows you to resume any past Claude Code session for the current project. All context and conversation history is preserved, eliminating the need to re-explain your project after restarts or VS Code refreshes.
When Claude starts hallucinating or performing poorly, it's often due to context window bloat. The /context command shows a visualization of what's consuming your context - system prompts, tools, past messages, etc. Use /clear to remove message history when needed.
"I hear a lot of people on X complain right now, oh, Claude's getting worse. No, that's a skill issue - they're just not managing their context well."
The /stats command reveals your Claude Code usage: tokens consumed, usage streaks, session counts, active days, peak hours, and model distribution. This helps track whether you're approaching plan limits and which models you prefer.
When running multiple sessions simultaneously, use /rename [name] to label each session. This makes it easy to identify specific features or tasks when using claude --res to resume later.
Press Escape twice to access rewind options. You can restore both code and conversation to any previous point, making experimentation risk-free. Choose whether to restore just the code, just the conversation, or both.
"Ever since they released this double escape feature, it's given me a lot more freedom to try new things."
Add "ultrathink" anywhere in your prompt to force Claude Code to work extra hard on complex problems. The word displays in rainbow colors when recognized. Best used for complex tasks, bug solving, and building core application features - not for every prompt due to increased token usage.
Instead of cluttering Claude rules, use the hashtag (#) to enter memory mode and add persistent instructions. Choose between user memory (global) or project memory (project-specific). Example: "# always use the design skill when changing the UI"
For experienced users, claude --dangerously-skip-permissions enables autonomous operation without approval prompts. This allows long-running tasks to complete without intervention, treating Claude as an independent employee.
"Claude for me hasn't made mistakes in months... Opus 4.5 has gotten so smart."
The /plugins command opens an app-store-like experience for Claude Code with skills, plugins, and MCPs available for one-click installation. Recommended: install the front-end design skill for dramatically better UI generation. Also available: GitHub, Figma, Linear, Notion integrations.
Press Ctrl+S to stash your current prompt, allowing you to write and send another prompt first. After sending, your stashed prompt automatically restores to the input line. No more holding backspace to clear long prompts.
"This is going to be the most important video you watch this month. I promise you whether you're a beginner or a pro, you will be using Claude Code completely differently after these 10 updates."
"The design skill is so freaking good. It's going to make your app look so much better."
"I like to give Claude long-running tasks where I can hit enter and walk away and I don't have to keep hitting accept over and over again."
| Command | Description |
|---|---|
claude --res | Resume past sessions |
/context | Visualize context window usage |
/clear | Clear message history |
/stats | View usage statistics |
/rename [name] | Name current session |
Escape Escape | Access rewind options |
ultrathink | Enable deep thinking mode (in prompt) |
# [memory] | Add custom memory |
claude --dangerously-skip-permissions | YOLO mode (skip approvals) |
/plugins | Open plugin store |
Ctrl+S | Stash current prompt |