agentscope

🥇Gold

AgentScope enables developers to build LLM applications using agent-oriented programming. It supports multi-agent systems, multi-modal interactions, and integrates with Claude. Ideal for creating complex chatbots and automation workflows.

23,3781,4160Updated Today

Intermediate30min to implementautomation

Saves ~100 min per use

Quick InstallView Source

git clone https://github.com/agentscope-ai/agentscope.git

Works with:

Claude

Overview

About This Skill

https://doc.agentscope.io/

How to Use

["Install AgentScope: Run `pip install agentscope` and verify the installation with `python -c \"import agentscope; print(agentscope.__version__)\"`.","Define agent roles: Create a Python script where each agent (e.g., Intake, Research, Resolution) is initialized with a specific function and tools (e.g., database access, email APIs). Use `agentscope.Agent` to structure their behavior.","Set up communication: Configure message passing between agents using AgentScope’s built-in `Message` class. For example, use `Message(sender='intake_agent', receiver='research_agent', content='Order #4567 delayed')` to trigger workflows.","Test and iterate: Run the system in a sandbox environment with mock data. Use AgentScope’s logging features to debug interactions and refine agent logic based on edge cases.","Deploy: Integrate with your existing tools (e.g., CRM, Slack) using AgentScope’s connectors. Monitor performance with built-in metrics and adjust agent prompts or tools as needed."]

Use Cases

Create a responsive chatbot that can handle customer inquiries and provide real-time support.

Develop a voice assistant capable of executing commands and providing information through natural language.

Implement a multi-agent system for complex workflows, allowing agents to collaborate and share information seamlessly.

Automate data processing tasks using specialized agents that can handle various data formats and sources.

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/agentscope-ai/agentscope

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Use AgentScope to design a multi-agent system for [TASK]. Define [AGENT_ROLES] and their interactions. Specify [INPUT_DATA] and [OUTPUT_FORMAT]. Ensure the agents coordinate to achieve [GOAL].

Example Output

In a customer support automation workflow, we deployed a 3-agent system using AgentScope:

1. **Intake Agent**: Received customer queries via email and Slack, extracting key details like order numbers and issue types. For example, it processed a query about a delayed shipment for Order #4567, categorizing it as a 'shipping delay' with high priority.

2. **Research Agent**: Queried the company’s CRM and logistics database to gather relevant data. It found that Order #4567 was shipped on June 10th but delayed due to a regional carrier issue. The agent also pulled the customer’s contact preferences and past interactions.

3. **Resolution Agent**: Drafted a personalized response to the customer, apologizing for the delay and offering a 15% discount on their next purchase. It scheduled the response to be sent via the customer’s preferred channel (email) and logged the interaction in the CRM.

The system reduced average response time from 4 hours to 20 minutes and improved customer satisfaction scores by 22% in pilot testing. The agents communicated asynchronously, with the Research Agent triggering the Resolution Agent only after confirming the shipment status.