This guide explains how to integrate the AgentScope Java framework documentation with AI-powered code editors.

• Cursor - AI-first code editor
• Windsurf - AI coding assistant
• GitHub Copilot with custom documentation
• Continue in VS Code
• Other editors supporting context documentation

1. Open Cursor Settings (⌘/Ctrl + ,)
2. Navigate to Features → Docs
3. Click "+ Add new Doc"
4. Paste the documentation URL:

   https://raw.githubusercontent.com/agentscope-ai/agentscope-java/main/docs/llm/agentscope-llm-guide.md 

5. Click "Add"

When writing code, load the documentation by:

1. Type @ in the chat
2. Select "Docs"
3. Choose "AgentScope Java" from the list
4. Ask your question or request code generation

@docs Create a ReActAgent with tools for weather and calculation @docs Show me how to implement streaming responses @docs Set up RAG with local documents @docs Configure long-term memory with Mem0 

1. Open Windsurf settings
2. Go to AI → Context
3. Add documentation file path or URL
4. Save configuration

Using AgentScope Java framework, create an agent with tool calling capability 

1. Create .copilot/docs/ directory in your project root
2. Copy agentscope-llm-guide.md to this directory
3. Copilot will automatically index it

Add a reference to the guide in your repository's Copilot instructions.

Install from VS Code marketplace: Continue

Edit .continue/config.json:

{ "docs": [ { "title": "AgentScope Java", "startUrl": "https://raw.githubusercontent.com/agentscope-ai/agentscope-java/main/docs/llm/agentscope-llm-guide.md" } ] }

Type @docs AgentScope Java in Continue chat to load context.

If you prefer local access:

1. Clone the repository:

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

2. Reference the local file in your IDE settings:

   file:///path/to/agentscope-java/docs/llm/agentscope-llm-guide.md 

Copy the guide to your project:

mkdir -p .copilot/docs curl -o .copilot/docs/agentscope-guide.md \ https://raw.githubusercontent.com/agentscope-ai/agentscope-java/main/docs/llm/agentscope-llm-guide.md

❌ Bad: "Create an agent" ✅ Good: "Create a ReActAgent with DashScope model, weather tool, and streaming enabled"

@docs Create an agent using: - Model: DashScope qwen3-max - Memory: AutoContextMemory with compression - Tools: weather and calculation - Mode: streaming 

Start simple, then add features:

1. @docs Create basic ReActAgent 2. @docs Add tool calling capability 3. @docs Enable streaming responses 4. @docs Add long-term memory 

@docs Explain how the ReAct loop works @docs What's the difference between AGENTIC and GENERIC RAG mode? @docs Show best practices for tool design 

1. @docs Show minimal AgentScope Java setup with Maven 2. @docs Create a basic chat agent with DashScope 3. @docs Add tool calling for web search and calculation 4. @docs Implement streaming responses 

@docs I have a ReActAgent. Show me how to: 1. Add local knowledge base 2. Configure document chunking 3. Use AGENTIC mode for retrieval 

@docs Create a multi-agent system where: - Agent A generates content - Agent B reviews it - Agent C makes final edits Use SequentialPipeline 

@docs Show production-ready agent with: - Error handling - Session persistence - Logging hooks - Resource cleanup 

Cursor:

• Check internet connection for URL-based docs
• Verify URL is correct
• Try re-adding the documentation

Local Files:

• Verify file path is correct
• Check file permissions
• Ensure file is readable

1. Explicitly reference with @docs

   • Type @ and select documentation

2. Include framework name in prompt

   • "Using AgentScope Java framework..."

3. Be specific about components

   • Mention specific classes like "ReActAgent", "Toolkit"

1. Check imports

   • Ensure all imports use io.agentscope.core.*

2. Verify Builder pattern usage

   • All objects use .builder()...build()

3. Review reactive programming

   • Proper use of Mono/Flux and .block()

I'm building a customer service chatbot that needs: - Multi-turn conversations with memory - Access to FAQ knowledge base - Ability to search products - Session persistence @docs Show me the agent setup 

@docs Explain when to use AutoContextMemory vs InMemoryMemory @docs What are the tradeoffs between RAG modes? 

@docs Show complete working example of agent with: - Error handling - Streaming - Multiple tools - Logging 

@docs Improve this agent code to add: - Better error handling - Resource cleanup - Tool execution logging 

The LLM guide is maintained in the AgentScope Java repository:

• Location: docs/llm/agentscope-llm-guide.md
• Updates: Synchronized with framework releases
• Issues: Report via GitHub Issues

To get the latest version:

curl -O https://raw.githubusercontent.com/agentscope-ai/agentscope-java/main/docs/llm/agentscope-llm-guide.md

Help improve the LLM guide:

1. Report Issues: Missing information, errors, unclear explanations
2. Suggest Examples: Common use cases that should be documented
3. Contribute: Submit PRs with improvements

See CONTRIBUTING.md for guidelines.

• Full Documentation: https://java.agentscope.io/
• GitHub Repository: https://github.com/agentscope-ai/agentscope-java
• Examples: Check agentscope-examples/ in the repository
• Discord Community: https://discord.gg/eYMpfnkG8h

Q: Does this work with all AI assistants? A: Most modern AI code assistants support custom documentation. The exact setup varies by tool.

Q: Can I modify the guide for my team? A: Yes! Fork the repository and customize. Consider contributing improvements back.

Q: How often is the guide updated? A: The guide is updated with each framework release and when significant features are added.

Q: What if the AI generates incorrect code? A: Always review generated code. Report persistent issues to improve the guide.

Q: Can I use this offline? A: Yes, download the guide locally and reference it in your IDE settings.

Happy coding with AgentScope Java! 🚀