Skip to main content

Next Steps

Now that you're signed in and familiar with the platform layout, here's a recommended path to get the most out of TraptureIQ. Follow these steps in order for the best experience.

For Tenant Admins (Workspace Owners)​

Step 1: Register Your First Agent​

Connect an AI agent to your workspace so you have something to work with.

How: Go to Admin → click + Register Agent → follow the Agent Registration guide.

Expected result: Your agent appears in the Agents list with a "Live" status. Your team can start chatting with it immediately.

Step 2: Start a Conversation​

Test your agent by having a real conversation.

How: Go to Agents → find your agent → click Chat with Agent → type a message and press Enter.

Expected result: Your agent responds in real-time with streaming text. You'll see token counts and latency metadata for each response.

Step 3: Explore Observability​

See what happened behind the scenes during your conversation.

How: After chatting, go to Traces to see the step-by-step execution, Logs to see the raw request/response data, and Sessions to see the session summary.

Expected result: You'll see exactly how your agent processed the request — which tools it called, how long each step took, and the full token breakdown.

Step 4: Invite Your Team​

Bring in colleagues so they can start using the platform too.

How: Go to Access Management → click Invite User → enter their email and select a role. See the Inviting Users guide.

Expected result: The user receives an invite email, and once they accept, they appear in your user list with the assigned role.

Step 5: Configure Security​

Set up AgentGuard to protect your agents from data leaks and malicious prompts.

How: Go to Settings → enable AgentGuard → then navigate to AgentGuard in the sidebar to configure firewall rules. See the AgentGuard guide.

Expected result: All agent conversations are now monitored in real-time. Blocked phrases, PII detection, and jailbreak prevention are active.

Step 6: Evaluate Agent Quality​

Test how well your agent performs with structured evaluations.

How: Go to Eval → choose Custom Eval to test response quality, Security Eval to test for vulnerabilities, or Load Test to test performance under load. See the Evaluation guide.

Expected result: You get scored results showing how your agent performs across metrics like accuracy, safety, and latency.

Step 7: Test MCP Integrations (Optional)​

If your agents use Google Cloud MCP tools, validate the connections.

How: Go to MCP Debug → click + Add Connection → select a server (e.g., BigQuery) → authenticate and test tools. See the MCP Debug guide.

Expected result: You can browse available MCP tools and execute test calls to verify they return the expected data.

For Tenant Users (Team Members)​

Step 1: Find Your Agents​

Browse the agents your admin has assigned to you.

How: Go to Agents in the sidebar. You'll see all agents you have access to displayed as cards.

Step 2: Start Chatting​

Pick an agent and have a conversation.

How: Click Chat with Agent on any agent card. Type your message and press Enter.

Step 3: Explore Available Modules​

Check which sidebar sections your admin has enabled for you and explore them:

  • Traces — See how your agent processes requests step by step
  • Logs — View detailed request/response logs
  • Sessions — Review your conversation history
  • Eval — Run quality and security tests on agents
  • Prompts — Create and manage prompt templates
  • MCP Debug — Test MCP integrations
  • Analyser — Count tokens and estimate costs (always available)
  • Intelligence — View agent reasoning insights (always available)

Step 4: Need More Access?​

If you need access to a sidebar section that's not visible, contact your Tenant Admin and ask them to enable it for you in Access Management → Section Access.