Skip to content
Kindo Kindo Docs

First Chat Experience

This quickstart walks you through your first AI-powered conversation in Kindo. By the end, you will have selected a model, sent a query, and used tools to enhance your results.

Prerequisites

  • A Kindo account (SaaS or self-hosted)
  • Signed in to the Kindo Terminal (see Getting Started)

Start a Conversation

  1. Select a model — Click the model selector above the query bar. For your first chat, choose a General Purpose model.

  2. Type a query — Enter a question or instruction in natural language. For example:

    Summarize the OWASP Top 10 vulnerabilities for 2024 and suggest mitigations for each.

  3. Press Enter — The model processes your query and streams the response in the main work area.

  4. Review the response — Kindo displays the full response with formatting. You can copy, share, or continue the conversation with follow-up questions.

Attach Files

You can provide additional context by attaching files to your query:

  1. Click the + button next to the query bar.
  2. Select one or more files from your machine.
  3. Use For All Files to apply your query to each file or to the collection as a whole.

File attachments are useful for tasks like analyzing log files, reviewing configurations, or summarizing documents.

Use Tools

Tools connect Kindo to your systems of record, extending what the model can access beyond its training data. Kindo supports a wide range of tool servers — from security platforms like CrowdStrike and SentinelOne to collaboration tools like Jira, Slack, and GitHub.

Use the Tool Policies panel to browse and manage available tool servers. Each tool server can be toggled to Active, Off, or Connect for first-time setup. Active tools are automatically available to the model during your conversation.

Tips

  • Be specific — The more context you provide, the better the response. Include relevant details about your environment, constraints, and desired format.
  • Iterate — Follow up with clarifying questions. Kindo maintains conversation context across messages.
  • Try different models — Switch between General Purpose, Reasoning, and Cybersecurity models to compare results for different tasks.

Next Steps