Your First Agent
Step-by-step guide to creating your first agent and achieving your first success moment
Your First Agent
Create and chat with your first agent in under five minutes. This guide walks you from the Agents page to your first successful conversation.
Navigation
- Parent: Getting Started
- Previous: Creating Your Account
- Next: Connecting Splunk
- Related: Understanding Agents | Purpose-Built Agents
Step 1: Navigate to Agents
From the dashboard, open the Agents section in the sidebar or main navigation. Click Create new agent to start.
Step 2: Pick a Template
When you create an agent, you choose a template. Templates pre-configure the agent with a role, system prompt, and suggested tools. For your first run, pick Search Ninja — a purpose-built agent for SPL generation, optimization, and best practices. Search Ninja helps you write and refine Splunk search queries.
Step 3: Configure the Model
Select the AI model for your agent. You can choose from OpenAI, Anthropic, Google, or xAI models depending on your plan and preferences. The template suggests a default; you can change it. Save the configuration when you're done.
Step 4: Start Chatting
Open your new agent and start a conversation. Type a question like "How do I search for failed logins in the last 24 hours?" or "Write an SPL query to find top 10 source IPs by event count." The agent responds with SPL, explanations, and suggestions. You've reached your first success moment when you get a useful, actionable response.
First Success Moment
Your first success is when the agent returns SPL or guidance you can use. If you've connected Splunk (see Connecting Splunk), the agent can run queries against your environment and show real results. Without Splunk connected, the agent still generates SPL and best practices you can copy into Splunk manually.
What's Next
Try different prompts, explore other purpose-built agents like Splunk Sensei for learning or BOTS Hunter for security investigations, or connect Splunk for context-aware assistance. See Understanding Agents for more on agent types and Chatting with Agents for conversation tips.