Skip to main content

Menu

Home
GitHub(在新标签页打开)

Home › Welcome

CrewAI Documentation

Documentation › Get Started

Introduction
Build with AI
Skills
Installation
Quickstart

Documentation › Guides › Strategy

Evaluating Use Cases for CrewAI

Documentation › Guides › Agents

Crafting Effective Agents

Documentation › Guides › Crews

Build Your First Crew

Documentation › Guides › Flows

Build Your First Flow
Mastering Flow State Management
Conversational Flows
Migrating from inputs.id to restore_from_state_id

Documentation › Guides › Frontend

Frontend Overview
Conversational Flows
Generative UI
Tool-Based Generative UI
Agentic Generative UI
Reasoning
A2UI
Human-in-the-Loop
Shared State
Frontend Actions
Predictive State Updates
Channels

Documentation › Guides › Tools

Publish Custom Tools

Documentation › Guides › Coding Tools

Coding Tools
Build with AI

Documentation › Guides › Advanced

Customizing Prompts
Fingerprinting

Documentation › Guides › Migration

Moving from LangGraph to CrewAI: A Practical Guide for Engineers
Upgrading CrewAI

Documentation › Core Concepts

Agents
Agent Capabilities
Tasks
Crews
Flows
Streaming
Production Architecture
Knowledge
Skills
LLMs
Files
Processes
Collaboration
Training
Memory
Reasoning
Planning
Testing
CLI
Tools
Event Listeners
Checkpointing

Documentation › MCP Integration

MCP Servers as Tools in CrewAI
MCP DSL Integration
Stdio Transport
SSE Transport
Streamable HTTP Transport
Connecting to Multiple MCP Servers
MCP Security Considerations

Documentation › Tools

Tools Overview

Documentation › Tools › File & Document

Overview
File Read
File Write
PDF RAG Search
DOCX RAG Search
MDX RAG Search
XML RAG Search
TXT RAG Search
JSON RAG Search
CSV RAG Search
Directory RAG Search
Directory Read
OCR Tool
PDF Text Writing Tool

Documentation › Tools › Web Scraping & Browsing

Overview
Scrape Website
Scrape Element From Website Tool
Scrapfly Scrape Website Tool
Selenium Scraper
Scrapegraph Scrape Tool
Spider Scraper
Browserbase Web Loader
Hyperbrowser Load Tool
Stagehand Tool
Firecrawl Crawl Website
Firecrawl Scrape Website
Oxylabs Scrapers
Bright Data Tools
You.com Content Extraction Tool

Documentation › Tools › Search & Research

Overview
Google Serper Search
Brave Search Tools
Exa Search Tool
Linkup Search Tool
Github Search
Website RAG Search
Code Docs RAG Search
YouTube Channel RAG Search
YouTube Video RAG Search
Tavily Search Tool
Tavily Extractor Tool
Tavily Research Tool
Arxiv Paper Tool
SerpApi Google Search Tool
SerpApi Google Shopping Tool
Databricks SQL Query Tool
You.com Search & Research Tools

Documentation › Tools › Database & Data

Overview
MySQL RAG Search
PG RAG Search
Snowflake Search Tool
NL2SQL Tool
Qdrant Vector Search Tool
Weaviate Vector Search
MongoDB Vector Search Tool
SingleStore Search Tool
Db2 Vector Search Tool

Documentation › Tools › AI & Machine Learning

Overview
DALL-E Tool
Vision Tool
AI Mind Tool
LlamaIndex Tool
LangChain Tool
RAG Tool
Code Interpreter
Daytona Sandbox Tools
E2B Sandbox Tools

Documentation › Tools › Cloud & Storage

Overview
S3 Reader Tool
S3 Writer Tool
Bedrock Knowledge Base Retriever

Documentation › Tools › Integrations

Overview
Bedrock Invoke Agent Tool
CrewAI Run Automation Tool
Merge Agent Handler Tool

Documentation › Tools › Automation

Overview
Apify Actors
Composio Tool
MultiOn Tool
Wait Tool
Zapier Actions Tool

Documentation › Observability

CrewAI Tracing
Overview
Arize Phoenix
Braintrust
Datadog Integration
Galileo
LangDB Integration
Langfuse Integration
Langtrace Integration
Maxim Integration
MLflow Integration
Neatlogs Integration
OpenLIT Integration
Opik Integration
Patronus AI Evaluation
Portkey Integration
Weave Integration
TrueFoundry Integration

Documentation › Learn

Overview
Strategic LLM Selection Guide
Conditional Tasks
Coding Agents
Create Custom Tools
Custom LLM Implementation
Custom Manager Agent
Customize Agents
Image Generation with DALL-E
Force Tool Output as Result
Hierarchical Process
Human Input on Execution
Human-in-the-Loop (HITL) Workflows
Human Feedback in Flows
Kickoff Crew Asynchronously
Kickoff Crew for Each
Streaming Runtime Contract
Consuming Streams
Connect to any LLM
Using CrewAI Without LiteLLM
Using Multimodal Agents
Replay Tasks from Latest Crew Kickoff
Sequential Processes
Using Annotations in crew.py

Documentation › Learn › Execution Hooks

Execution Hooks
LLM Call Hooks
Tool Call Hooks
Execution Boundary Hooks
Step Hooks
Before and After Kickoff Hooks

Documentation › Telemetry

Telemetry

API Reference › Getting Started

Introduction
GET /inputs
POST /kickoff
POST /resume
GET /status/{kickoff_id}

Examples

CrewAI Examples
CrewAI Cookbooks

Changelog › Release Notes

Changelog

本页还没有译文

本页 1 段还没译成简体中文，下面显示的是英文原文。

POST /kickoff

还没翻译译自英文（docs.crewai.com）本页 1 段都还没译

AI 翻译·查看英文原文

Adapted from docs.crewai.com/en/api-reference/kickoff by crewAI, Inc., licensed under MIT. Changes were made: this page is re-rendered from the original's source.

上一页GET /inputs 下一页POST /resume

译文 © 2026 fanyi.dev · 原文 © 2026 crewAI, Inc. (MIT)

加宽正文