Boost Your Productivity 10× with Claude Code Slash Commands

Overview

The author introduces Claude Code’s slash commands, which appear when you type / and cover session management, context control, parallel execution, and more. Using these commands correctly can save time and reduce token usage.

1. Session Management

/clear – Clears the current conversation to start fresh. Useful when switching to a new task; the previous context can be restored later with /resume.

/compact – Compresses the conversation to free up context space. When the token window reaches ~95%, Claude auto‑compresses, but the author recommends running it manually after each logical step (e.g., after debugging a bug). Example:

/compact 重点保留 API 接口的设计决策和参数定义

.

/resume – Restores a cleared or previous conversation. You can select a specific session from the selector or specify a name/ID, e.g., /resume 昨天的重构任务.

/branch and /fork – Create a new branch from the current dialogue to explore alternative solutions without affecting the original flow. Example: /branch 试试用 Redis 替代内存缓存的方案.

/rewind – Reverts the conversation and code to an earlier checkpoint, similar to git reset. You choose the target node from a popup.

2. Information & Diagnosis

/usage – Shows token consumption and estimated cost. Important to avoid unexpected charges; note that third‑party models may report inaccurate costs.

/context – Visualizes current context usage as a colored grid. Adding all expands details. The author pairs it with /compact to decide when to compress.

/diff – Opens an interactive diff viewer for uncommitted changes. The author suggests using a GUI diff tool for large changes.

/status – Opens the settings tab showing version, model, account, and connection status. Can be invoked while Claude is busy.

/help – Lists all available commands with brief descriptions.

/insights – Generates an analysis report of your Claude Code usage, including frequent directories, interaction patterns, and problem areas.

3. Model & Mode Control

/plan – Enters planning mode, prompting Claude to produce an execution plan before writing code. Example: /plan 重构整个项目，增加用户系统和 JWT 认证.

/goal – Sets a completion condition so Claude works autonomously until the condition is met. The author warns to write verifiable conditions (e.g., test exit code 0) and to add a safety limit, e.g.,

/goal 把所有 API 调用迁移到 v2 格式，直到测试通过，如果 20 轮还没搞定就停下来

.

/model , /effort , /fast – Adjust the underlying AI model, reasoning effort (low to max), and enable fast mode (≈2.5× speed, higher token price).

4. Configuration & Extensions

/config – Opens the settings UI for theme, model, output style, etc.

/mcp , /skills , /plugin – Manage MCP server connections, custom skill lists, and plugin bundles. These are rarely needed after initial setup.

5. Code Review

/review – Reviews the current branch’s pull request, automatically launching parallel agents to find bugs.

/simplify – Reviews recent changes and directly applies fixes for code reuse, quality, and efficiency. You can focus the fix, e.g., /simplify 重点关注内存效率.

/ultrareview – A deeper cloud‑based multi‑agent review that consumes extra usage credits (three free runs for Pro/Max users).

6. Agents & Parallel Execution

/agents – Shows and configures sub‑agents; most users let Claude manage them automatically.

/tasks – Lists background tasks in the current session; selecting a task shows its detailed output.

/background – Detaches the whole session to run in the background, freeing the terminal. You can add a final instruction, e.g.,

/background 编写单元测试并运行，执行失败则自动修复

.

/loop – Sets a periodic task, such as /loop 5m 检查项目前后端的部署状态. The author cautions about forgetting to stop loops, which can waste resources.

Conclusion

The author emphasizes that you don’t need to memorize every command; simply typing / shows the list, and describing your intent in Chinese lets Claude pick the appropriate command automatically. Mastering these slash commands can significantly reduce cognitive load in the AI‑assisted programming workflow.