> For the complete documentation index, see [llms.txt](https://ludicore-ai.gitbook.io/ludicore-ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://ludicore-ai.gitbook.io/ludicore-ai/achieve-superior-outcomes.md). # Achieve Superior Outcomes LudiCore is built to simplify and enhance your game development process. To fully leverage the power of our AI assistant, follow these best practices and techniques. Here’s how to improve the quality of your results and streamline your workflow for maximum efficiency. ## Divide Tasks into Manageable Steps Instead of trying to tackle an entire project in one large prompt, break it into smaller, actionable steps. Start by outlining the high-level tasks required to achieve your goal and approach them one at a time. **Example: Creating a First-Person Shooter Game** 1. **Build the Map:** Begin by asking the AI to generate the environment for your game. 2. **Create the Player Object:** Next, have the AI design and set up your player character. 3. **Implement Movement:** Finally, ask the AI to create a script for player movement. By focusing on individual tasks, the AI can produce more accurate and context-specific outputs without being overwhelmed by complexity. This step-by-step approach ensures a smoother workflow and better results. *** ## **Leveraging Context for Maximum Effectiveness** LudiCore's context feature lets you specify relevant game objects, scripts, and assets, helping the AI focus on what matters most. Providing targeted context ensures better and more accurate responses, so it’s a good habit to guide the AI with relevant information instead of leaving it to solve problems blindly. #### **Best Practices:** **Drag and Drop Context:**\ When working on a specific task, like creating an enemy controller, drag and drop the relevant object or script (e.g., enemy object details) into the context window. This helps the AI tailor its solutions to your specific enemy behavior instead of generating generic suggestions. **Avoid Overloading Context:**\ Providing too much context can reduce the quality of the response, as the AI may struggle to prioritize what’s important. Instead of adding every object and script, focus on including only the most relevant details for each prompt to achieve better outcomes. By thoughtfully using the context feature, you can guide LudiCore to deliver more precise and efficient solutions tailored to your project. ## Craft Clear and Detailed Prompts The quality of your input directly impacts the quality of the AI’s output. Crafting great prompts can be challenging, but following these tips will help you get the best results: #### **1. Be Specific and Descriptive** Provide clear details about what you want to achieve. Vague prompts lead to generic responses, while precise prompts enable the AI to tailor its solutions to your needs. * **Instead of asking:**\ \&#xNAN;*"Make my enemy kill the player."* * **Say:**\ \&#xNAN;*"Make the enemy periodically fire projectiles that home in on the player and kill them upon impact."* This ensures the AI understands your intent and avoids guessing, delivering a response designed for your specific requirements. *** #### **2. Use Supporting Information** Include relevant context to help the AI better frame its responses: * If you’re creating a metroidvania, let the AI know. * If you want the game to feel like Minecraft, include that detail. These background insights help the AI provide responses aligned with your vision. *** #### **Pro Tip:** Use `/chat` to generate an outline for your project or feature. Then: 1. Take the AI’s suggested steps and use each as a focused prompt. 2. Copy and paste step 1 into a new prompt, guide the AI through it, and refine the results. 3. Once you’re satisfied with that step, move on to step 2, and so on. By iterating this way, you can systematically build your project while maintaining clarity and control over each phase of development. *** ## Choose the Appropriate Chat Mode Each slash command is tailored for a specific purpose, helping you save time and streamline your workflow. For detailed information and best practices, visit ⚙️ Slash Commands. * **Use `/chat`** for general questions or high-level guidance. * **Use `/script`** to create, debug, refactor, or edit code directly. * **Use `/prototype`** to set up scenes, objects, or components within your game engine. Choosing the appropriate mode ensures you’re fully leveraging LudiCore’s capabilities for optimal efficiency and precision. *** ## Begin Separate Chats for Each Feature While it may seem convenient to keep everything in one chat, it’s more effective to start a new chat for each feature or task. **Why?** * It keeps the context window focused, ensuring the AI’s responses aren’t influenced by older, unrelated chat history. * It helps you maintain clarity and precision for each aspect of your project. **Recommendation:**\ Instead of one chat for the entire project, create individual chats for specific features, such as: * *“Player Movement”* * *“Enemy AI”* * *“Level Design”* This approach not only improves organization but also allows you to revisit successful features easily, providing a clear and structured history of your progress. ## Learn AI Models and Their Strengths LudiCore utilizes multiple AI models, each tailored for specific tasks. Understanding their strengths can help you achieve better results: * **Base Model (DeepSeek v3):** Ideal for code support and general problem-solving. * **GPT-4omini:** Reliable for prototyping and general-purpose problem-solving. * **Haiku:** Designed for speed and exceptional context awareness. * **Sonnet & GPT-4o:** Versatile all-rounders with a focus on scripting capabilities. Experimenting with these models will allow you to identify the best fit for your specific needs and maximize your workflow efficiency. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://ludicore-ai.gitbook.io/ludicore-ai/achieve-superior-outcomes.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.