菜单
Build AI agents with Google's Agent Development Kit (ADK) Python - an open-source toolkit for building, evaluating, and deploying AI agents. Features LlmAgent, workflow agents (sequential, parallel, loop), tool integration, multi-agent systems, and deployment to Vertex AI or Cloud Run.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Export Claude Code sessions to searchable markdown with QMD integration. Use when user says "sync sessions", "export sessions", "search sessions", "session history", or wants to backup/search past conversations. Don't use for live note-taking, meeting transcripts, or non-Claude-Code content.
Convert product briefs, websites, documents into structured product specs. Generates modular markdown files for landing pages, guides, marketing.
Generate step-by-step migration guides between tools, frameworks, or workflows. Outputs: Phase-by-phase migration plan, compatibility mapping, rollback strategy. Use when switching tools, upgrading frameworks, or adopting new workflows. Triggers: migrate from X to Y, switch to, upgrade from, moving from, /migrate
Generate complete project setup recipes combining Vibery templates, MCPs, and configurations. Outputs: Install commands, CLAUDE.md, folder structure, workflow setup. Use when starting new projects or adding major features to existing ones. Triggers: start new project, setup project, project blueprint, /blueprint
Transform conversations and ideas into structured technical specifications. Outputs: User stories, acceptance criteria, technical requirements, edge cases. Use when user wants to document requirements before coding. Triggers: write spec, create user stories, document requirements, /spec
Chain Vibery commands, skills, and agents into repeatable automated workflows. Outputs: Executable workflow files, hook configurations, CI/CD integration. Use when user has repetitive multi-step processes to automate. Triggers: create workflow, automate this process, chain these commands, /workflow
| name | google-adk-python |
| description | Build AI agents with Google's Agent Development Kit (ADK) Python - an open-source toolkit for building, evaluating, and deploying AI agents. Features LlmAgent, workflow agents (sequential, parallel, loop), tool integration, multi-agent systems, and deployment to Vertex AI or Cloud Run. |
| license | MIT |
Open-source, code-first toolkit for building, evaluating, and deploying AI agents.
pip install google-adk
LLM-powered agents with dynamic routing and adaptive behavior.
from google.adk.agents import LlmAgent
from google.adk.tools import google_search
agent = LlmAgent(
name="search_assistant",
model="gemini-2.5-flash",
instruction="You are a helpful assistant that searches the web.",
tools=[google_search]
)
Execute agents in defined order.
from google.adk.agents import SequentialAgent
workflow = SequentialAgent(
name="research_workflow",
agents=[researcher, summarizer, writer]
)
Run multiple agents concurrently.
from google.adk.agents import ParallelAgent
parallel = ParallelAgent(
name="parallel_research",
agents=[web_researcher, paper_researcher]
)
# Specialized agents
researcher = LlmAgent(
name="Researcher",
model="gemini-2.5-flash",
tools=[google_search]
)
writer = LlmAgent(
name="Writer",
model="gemini-2.5-flash",
)
# Coordinator
coordinator = LlmAgent(
name="Coordinator",
model="gemini-2.5-flash",
instruction="Delegate tasks to researcher and writer.",
sub_agents=[researcher, writer]
)
from google.adk.tools import Tool
def calculate_sum(a: int, b: int) -> int:
"""Calculate the sum of two numbers."""
return a + b
sum_tool = Tool.from_function(calculate_sum)