Skip to main content

Quick Start

Get up and running with OpsKat in under 5 minutes. This guide walks you through initial setup, adding your first server, and running your first AI-assisted operation.

First Launch​

Launch OpsKat from your applications menu or run the built binary directly. On first launch, the app initializes its SQLite database and creates the data directory automatically.

The main interface has a sidebar with:

  • Assets — Your infrastructure inventory
  • AI Agent — Natural-language operations
  • Settings — Configuration and preferences

Adding Your First Asset​

SSH Server​

  1. Click the + button in the asset sidebar.
  2. Select SSH as the asset type.
  3. Fill in the connection details:
    • Name — A display name (e.g., "Production Web Server")
    • Host — Hostname or IP address
    • Port — SSH port (default: 22)
    • Username — SSH login user
    • Auth Type — Choose Password or Key
    • Password / Private Key — Your credentials
  4. Click Save.

Credentials are encrypted with AES-256-GCM before being stored in the local database. The master key is resolved from explicit configuration, the OS keyring, or a protected data-directory key file.

Database (MySQL / PostgreSQL)​

  1. Click + and select Database.
  2. Choose a Driver — MySQL, PostgreSQL, SQL Server, or SQLite. SQLite uses a local or remote file source rather than a network host.
  3. Enter host, port, username, password, and default database name.
  4. Optionally select an SSH asset to tunnel through.
  5. Click Save.

Redis​

  1. Click + and select Redis.
  2. Enter host, port, and optional password.
  3. Optionally select an SSH asset to tunnel through.
  4. Click Save.

RDP Remote Desktop​

  1. Click + and select RDP.
  2. Enter the Windows host, username, password, and optional domain.
  3. Optionally route the connection through an SSH tunnel or SOCKS proxy.
  4. Leave clipboard synchronization enabled if you want to copy text and files between local and remote desktops.
  5. Test the connection and click Save.

See the RDP guide for session controls and clipboard behavior.

Object Storage​

  1. Click + and select Object Storage.
  2. Choose a convenience preset or the generic S3-compatible option, then confirm the endpoint and region.
  3. Enter the Access Key ID and Secret Access Key, or select a managed credential.
  4. Adjust path-style access or TLS settings when required by the service.
  5. Test the connection and click Save.

Open the asset to browse buckets, objects, and transfers. See the Object Storage guide.

Connecting to an SSH Terminal​

  1. Select an SSH asset from the sidebar.
  2. Click Connect or double-click the asset.
  3. An interactive terminal session opens in a new tab.

You can split the terminal pane, open multiple sessions, and use the built-in SFTP file browser to transfer files.

Setting Up the AI Agent​

The AI Agent lets you manage infrastructure through natural language. Before using it, configure a provider.

  1. Open the AI Agent tab — the setup wizard will appear on first use.
  2. Select OpenAI Compatible as the provider type.
  3. Enter your API endpoint and API key.
  4. Choose a model (click Fetch Models to auto-discover, or type manually — model parameters will auto-fill).
  5. Click Complete.

This works with any OpenAI-compatible API, including DeepSeek, Azure OpenAI, self-hosted models, etc.

Option 2: Anthropic API​

  1. In the setup wizard, select Anthropic as the provider type.
  2. Enter your API key.
  3. Choose a Claude model (e.g., claude-sonnet-4-20250514).
  4. Click Complete.

You can manage multiple providers and switch between them in Settings > AI.

Your First AI Conversation​

  1. Open the AI Agent tab.

  2. Type a natural-language instruction, for example:

    Show me the disk usage on my production server

  3. The AI will:

    • Call list_assets to find available servers
    • Call run_command with df -h on the target asset
    • Return the results in a readable format

  4. If the command matches a deny policy, it will be blocked. If it requires confirmation, you will see a prompt before execution.

The AI tool calls are recorded in the Audit Log, including available policy and approval decision context.

What's Next​

  • Asset Management — Organize assets into groups, import/export
  • SSH Terminal — Split pane, SFTP, jump hosts, port forwarding
  • RDP Remote Desktop — Embedded remote desktop, special keys, fullscreen, and clipboard sync
  • AI Agent — Provider configuration, tools, policy enforcement
  • Query Editor — SQL and data-management consoles
  • Object Storage — S3-compatible bucket and object management
  • Policy — Configure allow/deny rules
  • Audit — Review audit logs and approval workflows