# Prompt Guide **How do I create a good prompt?** 1. Be clear, detailed, and specific The more complete and precise you are in your prompt, the more likely you are to receive an output that adheres to your creative vision. 2. Define subtasks If you are working on a multi-step project, break up your request into discrete jobs to be done to help the agent stay on track. 3. Re-state goals during continued iterations AI projects work best when approached iteratively, refining and updating the process over time with each new generation. As your project evolves, re-state your initial project goals and context. **How can I specify which model the agent uses?** Include the exact model you’d like to use in your prompt. You can also explicitly ask the agent to store your preferences in its memory. **How do I reference a previous output in a new generation?** Use the @ command or select an output from your Library. **How do I efficiently change only one element of a generation?** Specificity is key to receiving a desired output. Be as precise as possible about the elements of an existing generation that you want to change and also those that you want to stay the same. You can learn more in our Prompt Guide. **Can I customize my agent’s personality?** There are no specific personality settings, but you can prompt an agent to behave a specific way and ask it to store that behavior in its memory for consistency across sessions. --- # Agent Instructions: 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://docs.glif.app/getting-started/prompt-guide.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.