> ## Documentation Index > Fetch the complete documentation index at: https://docs.lyzr.ai/llms.txt > Use this file to discover all available pages before exploring further. # Agent Building > How to build an agent on Lyzr Studio? ## Step-by-Step Guide to Building an Agent on Lyzr Studio **Step 1: Create a new agent**\ Navigate to **Agent Builder** from the sidebar menu and select **Create New Agent**.\\ **Step 2: Provide Name & Description**\ Enter a unique name and provide a detailed description of the agent’s purpose.\\ **Step 3: Choose the right LLM model**\ Choose an LLM provider from the list and select a suitable model. You can click on the "configure icon" to adjust parameters like temperature and Top P.\\ **Step 4: Provide Agent role and instructions**\ Define the Agent Role (e.g., Customer Support, Data Fetcher) and provide instructions to guide its behavior (e.g., "Craft professional, personalized emails based on user input"). Use the **Improve** button to enhance the role and instructions for optimal LLM performance (optional but recommended).\\ **Step 5: Add Tools (Optional)**\ Tools extend the agent’s capabilities by allowing specific actions (e.g., sending emails). Adding tools makes your agent more actionable and useful.\ [*Quick guide to add Tools*](https://docs.lyzr.ai/agent-studio/tools)\\ **Step 6: Provide Examples to guide your agent through a template (Optional)**\\ **Step 7: Add Knowledge base to your agent if required (Optional)**\ Knowledge bases store relevant content that helps agents answer better.\ [*Quick guide to add Knowledge base*](https://docs.lyzr.ai/agent-studio/knowledge-base) **Step 8: Select the right features for your agent (Optional)**\ Choosing appropriate features leads to better agent output.\ [*Quick guide to choose the right features*](https://docs.lyzr.ai/agent-studio/features)\\ **Step 9: Create an agent**\ Once ready, click **Create Agent** to save your agent and start testing it.\\ **Step 10: Test your agent**\ Interact with your agent by asking questions and fine-tune responses by updating the configuration as needed.\\ **Step 11: Deploy agent (Optional)**\ After testing, deploy your agent either by publishing it in the Agent Marketplace or using the JSON & API key in your product.\ [*Quick guide to publishing agent*](https://docs.lyzr.ai/agent-studio/publishing_deployment)\\