Chat
The Chat interface is your primary way to interact with Hublvu. Through natural conversation, you can diagnose issues, execute remediation, create automation, and access your organization's knowledge.
Getting Started
When you open Chat, you'll see the welcome panel with:
- A greeting and quick-start guidance
- Quick Action buttons for common tasks
- Suggestions based on your recent activity
Click in the message input at the bottom, type your request, and press Enter.
What You Can Do
Ask Questions
Get answers from your organization's knowledge base and connected systems:
- "What's our procedure for handling database failover?"
- "Find documentation about the authentication system"
- "What alerts fired in the last hour?"
The assistant searches your institutional knowledge and uses available tools to provide contextual answers.
Diagnose Issues
Describe a problem and let the AI help investigate:
- "Users are reporting slow response times on the API"
- "Check the status of the payment service"
- "Why is the database connection pool exhausted?"
The assistant can execute diagnostic tools, query systems, and guide you through troubleshooting.
Run Flows and Guides
Execute existing automation:
- "Run the database health check flow"
- "Execute the deployment guide for the payment service"
- "Start the network diagnostics flow"
The assistant finds matching flows or guides and starts execution, prompting for any required inputs.
Use Custom Agents
Custom agents extend the chat with specialized capabilities. When your team creates custom agents (often by wrapping guides), you can invoke them through conversation:
- "Rotate the database credentials"
- "Check network connectivity to the production servers"
- "Run the backup verification"
The assistant recognizes trigger phrases and executes the appropriate agent, which may run a guide or perform specialized tasks using configured tools.
Create Automation
Build new flows, guides, and insights through conversation:
Flows — Describe an incident scenario:
- "Create a flow for diagnosing database connection failures"
- "Build a runbook for high memory usage on web servers"
Guides — Describe a procedure:
- "Create a guide for deploying our API service"
- "Build a runbook for onboarding new team members"
Insights — Describe what you want to analyze:
- "Show me run outcomes over the past month"
- "Create a chart of response times by severity"
The AI generates structured content based on your description, available tools, and knowledge base.
Conversations
Starting New Conversations
Click New Chat in the sidebar to start a fresh conversation. Each conversation maintains its own context and history.
Continuing Conversations
Your recent conversations appear in the sidebar below the Workspace section. Click any conversation to continue where you left off.
Branching Conversations
From any message, you can create a branch to explore an alternative direction without losing your original conversation path. This is useful when you want to try different approaches to solving a problem.
Tool Execution
When the AI uses tools to gather information or take action, you'll see a tool execution panel showing:
- Which tools are running
- Real-time progress
- Inputs and outputs
Understanding Tool Results
Tool results appear inline in the conversation. The AI interprets results and explains what they mean in context. You can expand tool execution details to see raw inputs and outputs.
When Tools Need Approval
Some tools or actions may require your explicit approval before execution. The assistant will explain what it wants to do and ask for confirmation.
Quick Actions
Quick Actions are pre-configured prompts that appear in the welcome panel. They provide one-click access to common operations.
To use a Quick Action:
- Start a new conversation (or scroll to the welcome panel)
- Click a Quick Action button
- The action executes and results appear in the conversation
Quick Actions are configured by your team administrators to match your most common workflows. See Quick Actions for configuration details.
Tips for Effective Use
Include system names, error messages, or specific scenarios. "Check the API" is less effective than "Check the payment API for 500 errors in the last 15 minutes."
Mention relevant services, environments, or constraints. The more context you provide, the more targeted the response.
Use follow-up questions to refine results. After the AI generates a flow, you can say "Add a step to check disk space" or "Change the timeout to 30 seconds."
When the AI creates flows, guides, or insights, review them in the editor to fine-tune details before using in production.