Use Ocelot

Use this guide when

Use this guide to manage EnderDash resources with Ocelot chat.

Before you begin

Before you start, find out:

• whether the task applies to one server or the whole organization
• which Ocelot tools are enabled in that organization
• whether host, shell, container, or database tools are enabled for the target
• whether Ocelot needs approval before it acts

If you do not know which tools are available, read Configure Ocelot Guardrails. You can also ask an admin.

Open Ocelot

Use Open Ocelot in the dashboard header. On a server page, start with that server selected. The Ocelot sidebar item under Manage opens organization settings for owners and admins.

Start with the right scope

• For a task on one runtime, select its server.
• For a task across servers, select the organization.

A smaller scope gives Ocelot less unrelated information.

Ask Ocelot to inspect before it changes anything

Use this sequence for operational tasks:

1. Ask Ocelot to inspect the current state.
2. Ask for a summary of the result.
3. Review the proposed action.
4. Approve or request the action.

These steps show you what Ocelot found before it changes the system.

Investigate one performance problem

On the affected server, ask:

Compare the answer with Statistics and Console. If Ocelot proposes a change, review its target and effect. After an approved action, verify the runtime result in a live panel.

If you want a finding saved, ask for a note that names the server, observed facts, and unresolved questions. Review the note before relying on it in a later session.

Keep requests concrete

A clear request names the item, action, and expected result.

Examples:

• Show me the last 100 console lines with errors only
• Find every file that mentions 25565
• Check whether the server is falling behind on TPS
• Show me the LuckPerms nodes for the moderator group
• Is Chunky still pregenerating the overworld?
• Check the Docker containers related to this server
• Look for recent Kubernetes events for this server workload

Review sensitive actions

Your organization can use these controls:

• high-impact actions can require approval
• some tool groups can be disabled entirely
• shell, container, database, and Kubernetes actions can use separate limits
• Emergency Stop can block Ocelot until an admin enables it again

If Ocelot refuses a request, an organization rule or permission usually blocks the action.

Save useful findings

Ask Ocelot to save important results as notes. Use notes for:

• recurring operational facts
• known server problems
• troubleshooting results that your team can use later

If Ocelot stops early

Make sure that:

• the organization still has available Ocelot usage
• the required tool group is enabled
• you selected the correct server or organization

Admins can change these controls in Configure Ocelot Guardrails.