SDK 中的 Slash Commands

了解如何透過 SDK 使用 slash commands 來控制 Claude Code 會話

Slash commands 提供了一種方式來控制 Claude Code 會話，使用以 / 開頭的特殊命令。這些命令可以透過 SDK 發送，以執行諸如壓縮上下文、列出上下文使用情況或調用自訂命令等操作。只有在不需要互動式終端的情況下才能工作的命令才能透過 SDK 分派；system/init 訊息列出了您會話中可用的命令。

發現可用的 Slash Commands

Claude Agent SDK 在系統初始化訊息中提供有關可用 slash commands 的資訊。在您的會話開始時存取此資訊：

import { query } from "@anthropic-ai/claude-agent-sdk";

for await (const message of query({
prompt: "Hello Claude",
options: { maxTurns: 1 }
})) {
if (message.type === "system" && message.subtype === "init") {
console.log("Available slash commands:", message.slash_commands);
// 包括內建命令加上捆綁的技能，例如：
// ["clear", "compact", "context", "usage", "code-review", "verify", ...]
}
}

import asyncio
from claude_agent_sdk import query, ClaudeAgentOptions, SystemMessage


async def main():
async for message in query(prompt="Hello Claude", options=ClaudeAgentOptions(max_turns=1)):
if isinstance(message, SystemMessage) and message.subtype == "init":
print("Available slash commands:", message.data["slash_commands"])
# 包括內建命令加上捆綁的技能，例如：
# ["clear", "compact", "context", "usage", "code-review", "verify", ...]


asyncio.run(main())

發送 Slash Commands

透過在您的提示字串中包含 slash commands 來發送它們，就像常規文字一樣。作用於對話歷史記錄的命令，例如 /compact，需要先前的訊息才能運作，因此下面的範例首先提出一個問題，然後將命令作為後續訊息發送到同一對話：

import { query } from "@anthropic-ai/claude-agent-sdk";

// Build up conversation history first
try {
for await (const message of query({
prompt: "What does the README in this directory cover?",
options: { maxTurns: 2 }
})) {
if (message.type === "result" && message.subtype === "success") {
console.log(message.result);
}
}
} catch (error) {
// A single-shot query() throws after yielding an error result,
// so the follow-up query below still runs.
console.error(`Session ended with an error: ${error}`);
}

// Send a slash command as a follow-up to the same conversation
for await (const message of query({
prompt: "/compact",
options: { continue: true, maxTurns: 1 }
})) {
if (message.type === "result") {
console.log("Command executed, result subtype:", message.subtype);
// Example output: Command executed, result subtype: success
}
}

import asyncio
from claude_agent_sdk import query, ClaudeAgentOptions, ResultMessage


async def main():
# Build up conversation history first
try:
async for message in query(
prompt="What does the README in this directory cover?",
options=ClaudeAgentOptions(max_turns=2),
):
if isinstance(message, ResultMessage) and message.subtype == "success":
print(message.result)
except Exception as error:
# A single-shot query() raises after yielding an error result,
# so the follow-up query below still runs.
print(f"Session ended with an error: {error}")

# Send a slash command as a follow-up to the same conversation
async for message in query(
prompt="/compact",
options=ClaudeAgentOptions(continue_conversation=True, max_turns=1),
):
if isinstance(message, ResultMessage):
print("Command executed, result subtype:", message.subtype)
# Example output: Command executed, result subtype: success


asyncio.run(main())

常見的 Slash Commands

`/compact` - 壓縮對話歷史

/compact 命令透過總結較舊的訊息同時保留重要上下文來減少您的對話歷史的大小。壓縮需要至少有兩次先前交換的現有對話才能進行總結。此範例首先進行對話，然後壓縮它並讀取報告結果的 compact_boundary 系統訊息：

import { query } from "@anthropic-ai/claude-agent-sdk";

// Compaction needs existing history, so have a conversation first
try {
for await (const message of query({
prompt: "Explain what this project does",
options: { maxTurns: 2 }
})) {
if (message.type === "result" && message.subtype === "success") {
console.log(message.result);
}
}
} catch (error) {
// A single-shot query() throws after yielding an error result,
// so the follow-up query below still runs.
console.error(`Session ended with an error: ${error}`);
}

// Compact the same conversation
for await (const message of query({
prompt: "/compact",
options: { continue: true, maxTurns: 1 }
})) {
if (message.type === "system" && message.subtype === "compact_boundary") {
console.log("Compaction completed");
console.log("Pre-compaction tokens:", message.compact_metadata.pre_tokens);
console.log("Trigger:", message.compact_metadata.trigger);
// Example output:
// Compaction completed
// Pre-compaction tokens: 1842
// Trigger: manual
}
}

import asyncio
from claude_agent_sdk import query, ClaudeAgentOptions, ResultMessage, SystemMessage


async def main():
# Compaction needs existing history, so have a conversation first
try:
async for message in query(
prompt="Explain what this project does",
options=ClaudeAgentOptions(max_turns=2),
):
if isinstance(message, ResultMessage) and message.subtype == "success":
print(message.result)
except Exception as error:
# A single-shot query() raises after yielding an error result,
# so the follow-up query below still runs.
print(f"Session ended with an error: {error}")

# Compact the same conversation
async for message in query(
prompt="/compact",
options=ClaudeAgentOptions(continue_conversation=True, max_turns=1),
):
if isinstance(message, SystemMessage) and message.subtype == "compact_boundary":
print("Compaction completed")
print("Pre-compaction tokens:", message.data["compact_metadata"]["pre_tokens"])
print("Trigger:", message.data["compact_metadata"]["trigger"])
# Example output:
# Compaction completed
# Pre-compaction tokens: 1842
# Trigger: manual


asyncio.run(main())

`/clear` - 重設對話上下文

/clear 命令將對話重設為空上下文，因此後續提示會從沒有先前對話歷史的狀態開始。先前的對話保存在磁碟上，可以透過將其會話 ID 傳遞給 resume 選項來返回。

這在串流輸入模式中很有用，您可以在單一連線上傳送多個提示。對於一次性的 query() 呼叫，每個呼叫已經開始時具有空上下文，因此傳送 /clear 沒有實際效果；請改為開始一個新的 query()。

建立自訂 Slash Commands

除了使用內建 slash commands 外，您還可以建立自己的自訂命令，這些命令可透過 SDK 使用。自訂命令定義為特定目錄中的 markdown 檔案，類似於子代理的配置方式。

檔案位置

自訂 slash commands 根據其範圍儲存在指定的目錄中：

專案命令：.claude/commands/ - 僅在目前專案中可用（舊版；建議使用 .claude/skills/）
個人命令：~/.claude/commands/ - 在您的所有專案中可用（舊版；建議使用 ~/.claude/skills/）

檔案格式

每個自訂命令都是一個 markdown 檔案，其中：

檔案名稱（不含 .md 副檔名）成為命令名稱
檔案內容定義命令的功能
可選的 YAML frontmatter 提供配置

基本範例

在您的專案中建立 .claude/commands 目錄（如果不存在），然後建立 .claude/commands/refactor.md：

Refactor the selected code to improve readability and maintainability.
Focus on clean code principles and best practices.

這會建立 /refactor 命令，您可以透過 SDK 使用。

使用 Frontmatter

建立 .claude/commands/security-check.md：

---
allowed-tools: Read, Grep, Glob
description: Run security vulnerability scan
model: claude-opus-4-8
---

Analyze the codebase for security vulnerabilities including:
- SQL injection risks
- XSS vulnerabilities
- Exposed credentials
- Insecure configurations

在 SDK 中使用自訂命令

一旦在檔案系統中定義，自訂命令就會自動透過 SDK 可用：

import { query } from "@anthropic-ai/claude-agent-sdk";

// Use a custom command
try {
for await (const message of query({
prompt: "/refactor src/auth/login.ts",
options: { maxTurns: 3 }
})) {
if (message.type === "assistant") {
console.log("Refactoring suggestions:", message.message);
}
}
} catch (error) {
// A single-shot query() throws after yielding an error result,
// so the second query below still runs.
console.error(`Session ended with an error: ${error}`);
}

// Custom commands appear in the slash_commands list
for await (const message of query({
prompt: "Hello",
options: { maxTurns: 1 }
})) {
if (message.type === "system" && message.subtype === "init") {
console.log("Available commands:", message.slash_commands);
// Includes built-in commands plus bundled skills and your custom commands, for example:
// ["clear", "compact", "context", "usage", "code-review", "verify", "refactor", "security-check", ...]
}
}

import asyncio
from claude_agent_sdk import query, ClaudeAgentOptions, AssistantMessage, SystemMessage


async def main():
# Use a custom command
try:
async for message in query(
prompt="/refactor src/auth/login.py", options=ClaudeAgentOptions(max_turns=3)
):
if isinstance(message, AssistantMessage):
for block in message.content:
if hasattr(block, "text"):
print("Refactoring suggestions:", block.text)
except Exception as error:
# A single-shot query() raises after yielding an error result,
# so the second query below still runs.
print(f"Session ended with an error: {error}")

# Custom commands appear in the slash_commands list
async for message in query(prompt="Hello", options=ClaudeAgentOptions(max_turns=1)):
if isinstance(message, SystemMessage) and message.subtype == "init":
print("Available commands:", message.data["slash_commands"])
# Includes built-in commands plus bundled skills and your custom commands, for example:
# ["clear", "compact", "context", "usage", "code-review", "verify", "refactor", "security-check", ...]


asyncio.run(main())

進階功能

引數和佔位符

自訂命令支援使用佔位符的動態引數：

建立 .claude/commands/fix-issue.md：

---
argument-hint: [issue-number] [priority]
description: Fix a GitHub issue
---

Fix issue #$0 with priority $1.
Check the issue description and implement the necessary changes.

在 SDK 中使用：

import { query } from "@anthropic-ai/claude-agent-sdk";

// Pass arguments to custom command
for await (const message of query({
prompt: "/fix-issue 123 high",
options: { maxTurns: 5 }
})) {
// Command will process with $0="123" and $1="high"
if (message.type === "result" && message.subtype === "success") {
console.log("Issue fixed:", message.result);
}
}

import asyncio
from claude_agent_sdk import query, ClaudeAgentOptions, ResultMessage


async def main():
# Pass arguments to custom command
async for message in query(prompt="/fix-issue 123 high", options=ClaudeAgentOptions(max_turns=5)):
# Command will process with $0="123" and $1="high"
if isinstance(message, ResultMessage):
print("Issue fixed:", message.result)


asyncio.run(main())

Bash 命令執行

自訂命令可以執行 bash 命令並包含其輸出：

建立 .claude/commands/git-commit.md：

---
allowed-tools: Bash(git add *), Bash(git status *), Bash(git commit *)
description: Create a git commit
---

## Context

- Current status: !`git status`
- Current diff: !`git diff HEAD`

## Task

Create a git commit with appropriate message based on the changes.

檔案參考

使用 @ 前綴包含檔案內容：

建立 .claude/commands/review-config.md：

---
description: Review configuration files
---

Review the following configuration files for issues:
- Package config: @package.json
- TypeScript config: @tsconfig.json
- Environment config: @.env

Check for security issues, outdated dependencies, and misconfigurations.

使用命名空間組織

在子目錄中組織命令以獲得更好的結構：

.claude/commands/
├── frontend/
│   ├── component.md      # Creates /component (project:frontend)
│   └── style-check.md     # Creates /style-check (project:frontend)
├── backend/
│   ├── api-test.md        # Creates /api-test (project:backend)
│   └── db-migrate.md      # Creates /db-migrate (project:backend)
└── review.md              # Creates /review (project)

子目錄出現在命令描述中，但不會影響命令名稱本身。

實用範例

Pull Request 審查命令

建立 .claude/commands/review-pr.md：

---
allowed-tools: Read, Grep, Glob, Bash(git diff *)
description: Comprehensive code review
---

## Changed Files
!`git diff --name-only HEAD~1`

## Detailed Changes
!`git diff HEAD~1`

## Review Checklist

Review the above changes for:
1. Code quality and readability
2. Security vulnerabilities
3. Performance implications
4. Test coverage
5. Documentation completeness

Provide specific, actionable feedback organized by priority.

測試執行器命令

建立 .claude/commands/test.md：

---
allowed-tools: Bash, Read, Edit
argument-hint: [test-pattern]
description: Run tests with optional pattern
---

Run tests matching pattern: $ARGUMENTS

1. Detect the test framework (Jest, pytest, etc.)
2. Run tests with the provided pattern
3. If tests fail, analyze and fix them
4. Re-run to verify fixes

透過 SDK 使用這些命令：

import { query } from "@anthropic-ai/claude-agent-sdk";

// Run code review
try {
for await (const message of query({
prompt: "/review-pr",
options: { maxTurns: 3 }
})) {
// Process review feedback
}
} catch (error) {
// A single-shot query() throws after yielding an error result,
// so the second query below still runs.
console.error(`Session ended with an error: ${error}`);
}

// Run specific tests
for await (const message of query({
prompt: "/test auth",
options: { maxTurns: 5 }
})) {
// Handle test results
}

import asyncio
from claude_agent_sdk import query, ClaudeAgentOptions


async def main():
# Run code review
try:
async for message in query(prompt="/review-pr", options=ClaudeAgentOptions(max_turns=3)):
# Process review feedback
pass
except Exception as error:
# A single-shot query() raises after yielding an error result,
# so the second query below still runs.
print(f"Session ended with an error: {error}")

# Run specific tests
async for message in query(prompt="/test auth", options=ClaudeAgentOptions(max_turns=5)):
# Handle test results
pass


asyncio.run(main())

另請參閱

Slash Commands - 完整的 slash command 文件
SDK 中的子代理 - 子代理的類似檔案系統配置
TypeScript SDK 參考 - 完整的 API 文件
SDK 概述 - 一般 SDK 概念
CLI 參考 - 命令列介面

agent-sdk/slash-commands.md +141 −25

24 })) {24 })) {

25 if (message.type === "system" && message.subtype === "init") {25 if (message.type === "system" && message.subtype === "init") {

26 console.log("Available slash commands:", message.slash_commands);26 console.log("Available slash commands:", message.slash_commands);

~~27 // Example output: ["clear", "compact", "context", "usage"]~~27 // 包括內建命令加上捆綁的技能，例如：

28 // ["clear", "compact", "context", "usage", "code-review", "verify", ...]

28 }29 }

29 }30 }

30 ```31 ```

38 async for message in query(prompt="Hello Claude", options=ClaudeAgentOptions(max_turns=1)):39 async for message in query(prompt="Hello Claude", options=ClaudeAgentOptions(max_turns=1)):

39 if isinstance(message, SystemMessage) and message.subtype == "init":40 if isinstance(message, SystemMessage) and message.subtype == "init":

40 print("Available slash commands:", message.data["slash_commands"])41 print("Available slash commands:", message.data["slash_commands"])

~~41 # Example output: ["clear", "compact", "context", "usage"]~~42 # 包括內建命令加上捆綁的技能，例如：

43 # ["clear", "compact", "context", "usage", "code-review", "verify", ...]

42 44

43 45

44 asyncio.run(main())46 asyncio.run(main())

49 發送 Slash Commands51 發送 Slash Commands

50</h2>52</h2>

51 53

~~52透過在您的提示字串中包含 slash commands 來發送它們，就像常規文字一樣：~~54透過在您的提示字串中包含 slash commands 來發送它們，就像常規文字一樣。作用於對話歷史記錄的命令，例如 `/compact`，需要先前的訊息才能運作，因此下面的範例首先提出一個問題，然後將命令作為後續訊息發送到同一對話：

53 55

54<CodeGroup>56<CodeGroup>

55 ```typescript TypeScript theme={null}57 ```typescript TypeScript theme={null}

56 import { query } from "@anthropic-ai/claude-agent-sdk";58 import { query } from "@anthropic-ai/claude-agent-sdk";

57 59

~~58 // Send a slash command~~60 // Build up conversation history first

61 try {

59 for await (const message of query({62 for await (const message of query({

~~60 prompt: "/compact",~~63 prompt: "What does the README in this directory cover?",

~~61 options: { maxTurns: 1 }~~64 options: { maxTurns: 2 }

62 })) {65 })) {

63 if (message.type === "result" && message.subtype === "success") {66 if (message.type === "result" && message.subtype === "success") {

~~64 console.log("Command executed:", message.result);~~67 console.log(message.result);

68 }

69 }

70 } catch (error) {

71 // A single-shot query() throws after yielding an error result,

72 // so the follow-up query below still runs.

73 console.error(`Session ended with an error: ${error}`);

74 }

76 // Send a slash command as a follow-up to the same conversation

77 for await (const message of query({

78 prompt: "/compact",

79 options: { continue: true, maxTurns: 1 }

80 })) {

81 if (message.type === "result") {

82 console.log("Command executed, result subtype:", message.subtype);

83 // Example output: Command executed, result subtype: success

65 }84 }

66 }85 }

67 ```86 ```

72 91

73 92

74 async def main():93 async def main():

~~75 # Send a slash command~~94 # Build up conversation history first

~~76 async for message in query(prompt="/compact", options=ClaudeAgentOptions(max_turns=1)):~~95 try:

96 async for message in query(

97 prompt="What does the README in this directory cover?",

98 options=ClaudeAgentOptions(max_turns=2),

99 ):

100 if isinstance(message, ResultMessage) and message.subtype == "success":

101 print(message.result)

102 except Exception as error:

103 # A single-shot query() raises after yielding an error result,

104 # so the follow-up query below still runs.

105 print(f"Session ended with an error: {error}")

106

107 # Send a slash command as a follow-up to the same conversation

108 async for message in query(

109 prompt="/compact",

110 options=ClaudeAgentOptions(continue_conversation=True, max_turns=1),

111 ):

77 if isinstance(message, ResultMessage):112 if isinstance(message, ResultMessage):

~~78 print("Command executed:", message.result)~~113 print("Command executed, result subtype:", message.subtype)

114 # Example output: Command executed, result subtype: success

79 115

80 116

81 asyncio.run(main())117 asyncio.run(main())

82 ```118 ```

83</CodeGroup>119</CodeGroup>

84 120

121<Note>

122 查詢可能以錯誤結果結束，例如當 `maxTurns` / `max_turns` 限制在工作完成前達到時。最終結果訊息則具有 `is_error: true` 和錯誤子類型（例如 `error_max_turns`）而不是 `success`。

123

124 在產生該最終結果訊息後，SDK 會拋出錯誤，因為 CLI 程序以非零代碼退出。

125

126 如果您的命令可能達到限制，請在 TypeScript 中使用 `try`/`catch` 或在 Python 中使用 `try`/`except` 包裝迴圈，如 [Single Message Input](/zh-TW/agent-sdk/streaming-vs-single-mode#single-message-input) 中所示，或設定 `maxTurns` 足夠高以完成工作。在 Python 中，捕捉 `Exception`：SDK 將錯誤結果表示為純 `Exception`。

127</Note>

128

85<h2 id="common-slash-commands">129<h2 id="common-slash-commands">

86 常見的 Slash Commands130 常見的 Slash Commands

87</h2>131</h2>

90 `/compact` - 壓縮對話歷史134 `/compact` - 壓縮對話歷史

91</h3>135</h3>

92 136

~~93`/compact` 命令透過總結較舊的訊息同時保留重要上下文來減少您的對話歷史的大小：~~137`/compact` 命令透過總結較舊的訊息同時保留重要上下文來減少您的對話歷史的大小。壓縮需要至少有兩次先前交換的現有對話才能進行總結。此範例首先進行對話，然後壓縮它並讀取報告結果的 `compact_boundary` 系統訊息：

94 138

95<CodeGroup>139<CodeGroup>

96 ```typescript TypeScript theme={null}140 ```typescript TypeScript theme={null}

97 import { query } from "@anthropic-ai/claude-agent-sdk";141 import { query } from "@anthropic-ai/claude-agent-sdk";

98 142

143 // Compaction needs existing history, so have a conversation first

144 try {

145 for await (const message of query({

146 prompt: "Explain what this project does",

147 options: { maxTurns: 2 }

148 })) {

149 if (message.type === "result" && message.subtype === "success") {

150 console.log(message.result);

151 }

152 }

153 } catch (error) {

154 // A single-shot query() throws after yielding an error result,

155 // so the follow-up query below still runs.

156 console.error(`Session ended with an error: ${error}`);

157 }

158

159 // Compact the same conversation

99 for await (const message of query({160 for await (const message of query({

100 prompt: "/compact",161 prompt: "/compact",

101 options: { maxTurns: 1 }162 options: { continue: true, maxTurns: 1 }

102 })) {163 })) {

103 if (message.type === "system" && message.subtype === "compact_boundary") {164 if (message.type === "system" && message.subtype === "compact_boundary") {

104 console.log("Compaction completed");165 console.log("Compaction completed");

105 console.log("Pre-compaction tokens:", message.compact_metadata.pre_tokens);166 console.log("Pre-compaction tokens:", message.compact_metadata.pre_tokens);

106 console.log("Trigger:", message.compact_metadata.trigger);167 console.log("Trigger:", message.compact_metadata.trigger);

168 // Example output:

169 // Compaction completed

170 // Pre-compaction tokens: 1842

171 // Trigger: manual

107 }172 }

108 }173 }

109 ```174 ```

110 175

111 ```python Python theme={null}176 ```python Python theme={null}

112 import asyncio177 import asyncio

113 from claude_agent_sdk import query, ClaudeAgentOptions, SystemMessage178 from claude_agent_sdk import query, ClaudeAgentOptions, ResultMessage, SystemMessage

114 179

115 180

116 async def main():181 async def main():

117 async for message in query(prompt="/compact", options=ClaudeAgentOptions(max_turns=1)):182 # Compaction needs existing history, so have a conversation first

183 try:

184 async for message in query(

185 prompt="Explain what this project does",

186 options=ClaudeAgentOptions(max_turns=2),

187 ):

188 if isinstance(message, ResultMessage) and message.subtype == "success":

189 print(message.result)

190 except Exception as error:

191 # A single-shot query() raises after yielding an error result,

192 # so the follow-up query below still runs.

193 print(f"Session ended with an error: {error}")

194

195 # Compact the same conversation

196 async for message in query(

197 prompt="/compact",

198 options=ClaudeAgentOptions(continue_conversation=True, max_turns=1),

199 ):

118 if isinstance(message, SystemMessage) and message.subtype == "compact_boundary":200 if isinstance(message, SystemMessage) and message.subtype == "compact_boundary":

119 print("Compaction completed")201 print("Compaction completed")

120 print("Pre-compaction tokens:", message.data["compact_metadata"]["pre_tokens"])202 print("Pre-compaction tokens:", message.data["compact_metadata"]["pre_tokens"])

121 print("Trigger:", message.data["compact_metadata"]["trigger"])203 print("Trigger:", message.data["compact_metadata"]["trigger"])

204 # Example output:

205 # Compaction completed

206 # Pre-compaction tokens: 1842

207 # Trigger: manual

122 208

123 209

124 asyncio.run(main())210 asyncio.run(main())

125 ```211 ```

126</CodeGroup>212</CodeGroup>

127 213

214<Note>

215 `compact_boundary` 訊息只在壓縮執行時才會到達。如果沒有任何內容可以總結，`/compact` 會報告原因而不是引發錯誤：執行仍然以 `success` 結果結束，不會發出 `compact_boundary` 訊息，結果文字會帶有訊息，例如在單一簡短交換後的 `Not enough messages to compact.`。全新的一次性 `query()` 呼叫開始時具有空上下文，因此請在具有先前輪次的會話中使用此模式，例如在[串流輸入模式](/zh-TW/agent-sdk/streaming-vs-single-mode)中或恢復會話時。

216</Note>

217

128<h3 id="/clear-reset-conversation-context">218<h3 id="/clear-reset-conversation-context">

129 `/clear` - 重設對話上下文219 `/clear` - 重設對話上下文

130</h3>220</h3>

170 基本範例260 基本範例

171</h4>261</h4>

172 262

173建立 `.claude/commands/refactor.md`：263在您的專案中建立 `.claude/commands` 目錄（如果不存在），然後建立 `.claude/commands/refactor.md`：

174 264

175```markdown theme={null}265```markdown theme={null}

176Refactor the selected code to improve readability and maintainability.266Refactor the selected code to improve readability and maintainability.

189---279---

190allowed-tools: Read, Grep, Glob280allowed-tools: Read, Grep, Glob

191description: Run security vulnerability scan281description: Run security vulnerability scan

192model: claude-opus-4-7282model: claude-opus-4-8

193---283---

194 284

195Analyze the codebase for security vulnerabilities including:285Analyze the codebase for security vulnerabilities including:

210 import { query } from "@anthropic-ai/claude-agent-sdk";300 import { query } from "@anthropic-ai/claude-agent-sdk";

211 301

212 // Use a custom command302 // Use a custom command

303 try {

213 for await (const message of query({304 for await (const message of query({

214 prompt: "/refactor src/auth/login.ts",305 prompt: "/refactor src/auth/login.ts",

215 options: { maxTurns: 3 }306 options: { maxTurns: 3 }

218 console.log("Refactoring suggestions:", message.message);309 console.log("Refactoring suggestions:", message.message);

219 }310 }

220 }311 }

312 } catch (error) {

313 // A single-shot query() throws after yielding an error result,

314 // so the second query below still runs.

315 console.error(`Session ended with an error: ${error}`);

316 }

221 317

222 // Custom commands appear in the slash_commands list318 // Custom commands appear in the slash_commands list

223 for await (const message of query({319 for await (const message of query({

225 options: { maxTurns: 1 }321 options: { maxTurns: 1 }

226 })) {322 })) {

227 if (message.type === "system" && message.subtype === "init") {323 if (message.type === "system" && message.subtype === "init") {

228 // Will include both built-in and custom commands

229 console.log("Available commands:", message.slash_commands);324 console.log("Available commands:", message.slash_commands);

230 // Example: ["clear", "compact", "context", "usage", "refactor", "security-check"]325 // Includes built-in commands plus bundled skills and your custom commands, for example:

326 // ["clear", "compact", "context", "usage", "code-review", "verify", "refactor", "security-check", ...]

231 }327 }

232 }328 }

233 ```329 ```

239 335

240 async def main():336 async def main():

241 # Use a custom command337 # Use a custom command

338 try:

242 async for message in query(339 async for message in query(

243 prompt="/refactor src/auth/login.py", options=ClaudeAgentOptions(max_turns=3)340 prompt="/refactor src/auth/login.py", options=ClaudeAgentOptions(max_turns=3)

244 ):341 ):

246 for block in message.content:343 for block in message.content:

247 if hasattr(block, "text"):344 if hasattr(block, "text"):

248 print("Refactoring suggestions:", block.text)345 print("Refactoring suggestions:", block.text)

346 except Exception as error:

347 # A single-shot query() raises after yielding an error result,

348 # so the second query below still runs.

349 print(f"Session ended with an error: {error}")

249 350

250 # Custom commands appear in the slash_commands list351 # Custom commands appear in the slash_commands list

251 async for message in query(prompt="Hello", options=ClaudeAgentOptions(max_turns=1)):352 async for message in query(prompt="Hello", options=ClaudeAgentOptions(max_turns=1)):

252 if isinstance(message, SystemMessage) and message.subtype == "init":353 if isinstance(message, SystemMessage) and message.subtype == "init":

253 # Will include both built-in and custom commands

254 print("Available commands:", message.data["slash_commands"])354 print("Available commands:", message.data["slash_commands"])

255 # Example: ["clear", "compact", "context", "usage", "refactor", "security-check"]355 # Includes built-in commands plus bundled skills and your custom commands, for example:

356 # ["clear", "compact", "context", "usage", "code-review", "verify", "refactor", "security-check", ...]

256 357

257 358

258 asyncio.run(main())359 asyncio.run(main())

384 實用範例485 實用範例

385</h3>486</h3>

386 487

387<h4 id="code-review-command">488<h4 id="pull-request-review-command">

388 程式碼審查命令489 Pull Request 審查命令

389</h4>490</h4>

390 491

391建立 `.claude/commands/code-review.md`：492建立 `.claude/commands/review-pr.md`：

392 493

393```markdown theme={null}494```markdown theme={null}

394---495---

414Provide specific, actionable feedback organized by priority.515Provide specific, actionable feedback organized by priority.

415```516```

416 517

518<Note>

519 Claude Code 包含捆綁的 `code-review` 和 `verify` skills。如果您以其中一個命名自訂命令，例如 `.claude/commands/code-review.md`，您的命令會遮蔽捆綁的 skill，而 `slash_commands` 列表會列出該名稱一次。

520</Note>

521

417<h4 id="test-runner-command">522<h4 id="test-runner-command">

418 測試執行器命令523 測試執行器命令

419</h4>524</h4>

442 import { query } from "@anthropic-ai/claude-agent-sdk";547 import { query } from "@anthropic-ai/claude-agent-sdk";

443 548

444 // Run code review549 // Run code review

550 try {

445 for await (const message of query({551 for await (const message of query({

446 prompt: "/code-review",552 prompt: "/review-pr",

447 options: { maxTurns: 3 }553 options: { maxTurns: 3 }

448 })) {554 })) {

449 // Process review feedback555 // Process review feedback

450 }556 }

557 } catch (error) {

558 // A single-shot query() throws after yielding an error result,

559 // so the second query below still runs.

560 console.error(`Session ended with an error: ${error}`);

561 }

451 562

452 // Run specific tests563 // Run specific tests

453 for await (const message of query({564 for await (const message of query({

465 576

466 async def main():577 async def main():

467 # Run code review578 # Run code review

468 async for message in query(prompt="/code-review", options=ClaudeAgentOptions(max_turns=3)):579 try:

580 async for message in query(prompt="/review-pr", options=ClaudeAgentOptions(max_turns=3)):

469 # Process review feedback581 # Process review feedback

470 pass582 pass

583 except Exception as error:

584 # A single-shot query() raises after yielding an error result,

585 # so the second query below still runs.

586 print(f"Session ended with an error: {error}")

471 587

472 # Run specific tests588 # Run specific tests

473 async for message in query(prompt="/test auth", options=ClaudeAgentOptions(max_turns=5)):589 async for message in query(prompt="/test auth", options=ClaudeAgentOptions(max_turns=5)):

agent-sdk/slash-commands.md 2026-06-29 23:02 UTC to 2026-06-30 23:02 UTC

SDK 中的 Slash Commands

發現可用的 Slash Commands

發送 Slash Commands

常見的 Slash Commands

`/compact` - 壓縮對話歷史

`/clear` - 重設對話上下文

建立自訂 Slash Commands

檔案位置

檔案格式

基本範例

使用 Frontmatter

在 SDK 中使用自訂命令

進階功能

引數和佔位符

Bash 命令執行

檔案參考

使用命名空間組織

實用範例

Pull Request 審查命令

測試執行器命令

另請參閱