create-workflow

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Vapi Workflow Creation

Vapi 工作流创建

Build structured conversation workflows with visual node-based flows. Workflows provide deterministic control over conversation steps, branching logic, and tool execution.

Setup: Ensure
VAPI_API_KEY
is set. See the
setup-api-key
skill if needed.

借助可视化节点流构建结构化对话工作流。工作流可为对话步骤、分支逻辑和工具执行提供确定性控制。

设置： 确保已配置
VAPI_API_KEY
。如有需要可参考
setup-api-key
skill。

When to Use Workflows vs Assistants

工作流与助手的适用场景对比

Feature	Assistant	Workflow
Simple conversations	Best choice	Over-engineered
Multi-step processes	Can work with good prompting	Best choice
Deterministic flow	Hard to guarantee	Built-in
Conditional branching	Prompt-dependent	Visual nodes
Complex state management	Difficult	Native support

功能	助手	工作流
简单对话	最优选择	过度设计
多步流程	提示词编写得当也可实现	最优选择
确定性流程	难以保证	原生支持
条件分支	依赖提示词实现	可视化节点配置
复杂状态管理	实现难度高	原生支持

Quick Start

快速开始

Workflows are best built in the Vapi Dashboard visual editor at https://dashboard.vapi.ai — but they can also be configured via API.

工作流最适合在 Vapi Dashboard 可视化编辑器中搭建，访问地址为 https://dashboard.vapi.ai —— 你也可以通过API进行配置。

Dashboard Workflow

仪表板创建工作流

Go to https://dashboard.vapi.ai
Navigate to Workflows
Click Create Workflow
Add nodes: Conversation, Tool, Condition, Handoff
Connect nodes to define the flow
Publish and attach to a phone number or call

访问 https://dashboard.vapi.ai
进入 Workflows 页面
点击 Create Workflow
添加节点：对话、工具、条件、转接
连接节点定义流程走向
发布后绑定到手机号或通话中使用

Using a Workflow in a Call

通话中使用工作流

bash

curl -X POST https://api.vapi.ai/call \
  -H "Authorization: Bearer $VAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "workflowId": "your-workflow-id",
    "phoneNumberId": "your-phone-number-id",
    "customer": {
      "number": "+11234567890"
    }
  }'

bash

curl -X POST https://api.vapi.ai/call \
  -H "Authorization: Bearer $VAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "workflowId": "your-workflow-id",
    "phoneNumberId": "your-phone-number-id",
    "customer": {
      "number": "+11234567890"
    }
  }'

Node Types

节点类型

Conversation Node

对话节点

The core building block — the assistant has a conversation within defined boundaries:

System prompt specific to this step
Model and voice configuration
Exit conditions that trigger transitions to other nodes
Variables to extract and pass between nodes

核心构建模块——助手在定义的边界内开展对话：

对应步骤专属的系统提示词
模型和音色配置
触发跳转到其他节点的退出条件
可提取并在节点间传递的变量

Tool Node

工具节点

Execute a tool (API call, function) and use the result in subsequent nodes.

执行工具（API调用、函数）并将结果用于后续节点。

Condition Node

条件节点

Branch the flow based on variables or conversation state.

基于变量或对话状态分支流程走向。

Handoff Node

转接节点

Transfer to another workflow, assistant, or phone number.

将对话转接到其他工作流、助手或手机号。

Workflow Patterns

工作流模式

Appointment Scheduling Flow

预约排期流

[Greeting] → [Collect Date] → [Check Availability (Tool)] → [Confirm Booking] → [Goodbye]
                                         ↓ (unavailable)
                                 [Suggest Alternatives] → [Confirm Booking]

[问候] → [收集日期] → [检查可用性（工具）] → [确认预约] → [结束] 
                                         ↓（无可用时段）
                                 [推荐替代时段] → [确认预约]

Lead Qualification Flow

线索资质判断流

[Introduction] → [Ask Budget] → [Ask Timeline] → [Qualify (Condition)]
                                                        ↓ (qualified)
                                                  [Schedule Demo]
                                                        ↓ (not qualified)
                                                  [Send Resources]

[自我介绍] → [询问预算] → [询问时间规划] → [资质判断（条件）]
                                                        ↓（符合要求）
                                                  [预约演示]
                                                        ↓（不符合要求）
                                                  [发送相关资源]

Support Triage Flow

客服分流流

[Greeting] → [Identify Issue (Condition)]
                  ↓ (billing)        ↓ (technical)        ↓ (other)
            [Billing Flow]    [Tech Support Flow]    [General Help]

[问候] → [识别问题类型（条件）]
                  ↓（账单问题）        ↓（技术问题）        ↓（其他问题）
            [账单处理流]    [技术支持流]    [常规帮助]

Attaching Workflows

绑定工作流

To a Phone Number

绑定到手机号

bash

curl -X PATCH https://api.vapi.ai/phone-number/{id} \
  -H "Authorization: Bearer $VAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "workflowId": "your-workflow-id"
  }'

bash

curl -X PATCH https://api.vapi.ai/phone-number/{id} \
  -H "Authorization: Bearer $VAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "workflowId": "your-workflow-id"
  }'

In an Outbound Call

外呼时使用

bash

curl -X POST https://api.vapi.ai/call \
  -H "Authorization: Bearer $VAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "workflowId": "your-workflow-id",
    "phoneNumberId": "your-phone-number-id",
    "customer": { "number": "+11234567890" }
  }'

bash

curl -X POST https://api.vapi.ai/call \
  -H "Authorization: Bearer $VAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "workflowId": "your-workflow-id",
    "phoneNumberId": "your-phone-number-id",
    "customer": { "number": "+11234567890" }
  }'

References

参考资料

Vapi Workflows Docs — Official guide
Workflow Examples — Common patterns

Vapi 工作流文档 —— 官方指南
工作流示例 —— 通用模式

Additional Resources

额外资源

This skills repository includes a Vapi documentation MCP server (

vapi-docs

) that gives your AI agent access to the full Vapi knowledge base. Use the

searchDocs

tool to look up anything beyond what this skill covers — advanced configuration, troubleshooting, SDK details, and more.

Auto-configured: If you cloned or installed these skills, the MCP server is already configured via

.mcp.json

(Claude Code),

.cursor/mcp.json

(Cursor), or

.vscode/mcp.json

(VS Code Copilot).

Manual setup: If your agent doesn't auto-detect the config, run:

bash

claude mcp add vapi-docs -- npx -y mcp-remote https://docs.vapi.ai/_mcp/server

See the README for full setup instructions across all supported agents.

本技能仓库包含一个 Vapi 文档 MCP 服务器（

vapi-docs

），可让你的AI Agent访问完整的Vapi知识库。如果需要查询本skill未覆盖的内容，可使用

searchDocs

工具查找高级配置、故障排查、SDK详情等信息。

自动配置： 如果你克隆或安装了这些技能，MCP服务器已经通过

.mcp.json

（Claude Code）、

.cursor/mcp.json

（Cursor）或

.vscode/mcp.json

（VS Code Copilot）完成配置。

手动设置： 如果你的Agent没有自动识别到配置，可执行：

bash

claude mcp add vapi-docs -- npx -y mcp-remote https://docs.vapi.ai/_mcp/server

查看README获取所有支持的Agent的完整设置说明。