EvoHub DocsEvoHub Docs Sign in

Get started with Catalog

Use with AI
View as MarkdownThis page as plain text, for pasting into an AI tool Open in ClaudeAsk Claude questions about this page Open in ChatGPTAsk ChatGPT questions about this page
Connect to Cursor / VS Code / ClaudeSearch and read these docs from your AI tool (MCP server)

MCP server URL

https://docs.evohub.io/mcp

Add to CursorAdd to VS Code

Claude Code

claude mcp add --transport http evohub-docs https://docs.evohub.io/mcp

Claude (claude.ai and Claude Desktop): Settings → Connectors → Add custom connector, and paste the URL above.

Claude Desktop — claude_desktop_config.json

{
  "mcpServers": {
    "evohub-docs": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://docs.evohub.io/mcp"
      ]
    }
  }
}

Cursor — ~/.cursor/mcp.json

{
  "mcpServers": {
    "evohub-docs": {
      "url": "https://docs.evohub.io/mcp"
    }
  }
}

VS Code — .vscode/mcp.json

{
  "servers": {
    "evohub-docs": {
      "type": "http",
      "url": "https://docs.evohub.io/mcp"
    }
  }
}

llms.txtllms-full.txt

This page takes you from an empty Catalog to services with owners and resources matched to them. You need permission to create environments and enroll agents; by default that is an Owner or Admin. See Catalog permissions.

1. Create an environment

An environment is one place you run software, such as a production cluster. Create one for each cluster, or each group of VMs, that you want to see.

Open Environments and agents

In the console, open Catalog, then Settings → Environments and agents, and select New environment.

Name it

Enter a Name such as eu-prod, pick its Kind (prod, staging, dev or other) and, if you like, a Description.

Create it

Select Create. The environment shows No agent reports here yet. until something reports into it.

2. Connect it

Catalog sees nothing on its own. Give it an agent, a cloud account, or both, for each environment.

Install an agent

The EvoHub agent runs in your Kubernetes cluster or on a Linux VM, reads what runs there with read-only access, and reports it. Secret values and environment variable values never leave your network.

Start from the environment

On the environment's card, select Install an agent, then Go to Agents. This opens Organization → Agents with that environment chosen; select New agent. Agents are managed there, not in Catalog, because one agent can serve several products.

Describe the agent

Enter a Name, choose Where it runs (Kubernetes cluster or Linux VM) and check the Environment.

Create the token

Select Create token. EvoHub creates a single-use enrollment token, valid for 15 minutes, and shows the install commands with it filled in.

Run the commands

Choose One command, or Keep it out of shell history to paste the token separately, then copy and run the commands. See Install on Kubernetes or Install on a VM for what they do.

The agent enrolls, sends a heartbeat every minute and a full sweep every five minutes. What it found appears in Catalog within a few minutes, and the agent shows as Reporting under Environments and agents and Organization → Agents. An agent that stops reporting shows Not reporting, and after ten minutes without a heartbeat what it reported is marked stale.

Connect a cloud account

For AWS, Azure, GCP or portvMind, EvoHub can read the account itself with read-only access you grant, every 15 minutes, with no agent to install. See Cloud accounts.

3. Declare your services

Open Services

Open Catalog → Services and select New service (or Declare your first service).

Fill in what you know

Enter the Name and the Slug. The slug is lowercase letters, digits and hyphens, for example payments-api, and cannot change later. Pick the Kind, Lifecycle, Owner team and Tier, and add the Repository, Docs and Runbook links. Only the name and slug are required.

Create it

Select Create service. Resources that already name this service in their annotation are matched to it at once.

4. Match what runs to your services

There are two ways a resource joins a service.

Annotate it. Add the evohub.io/component annotation (or label) with the service's slug to your manifests, for example on a Deployment:

metadata:
  annotations:
    evohub.io/component: payments-api

The next sweep binds it to that service without asking. Removing the annotation removes the binding at the next sweep.

Review the match queue. For everything without an annotation, Catalog proposes a service, for example when an application's name or app.kubernetes.io/name label is a service's slug. The Overview's Ownership panel shows how many suggestions are waiting and opens the Match queue, where you accept or reject each proposal. The Services entry in the navigation shows the same count. Accepting puts the application, and everything in it, in that service.

You can also open any application or resource and choose its service by hand. See Matching and drift.

5. Tie monitors and integrations to services

A service's SLA report is measured from what watches it. Set the Catalog service field:

  • on an Uptime monitor, when you create or edit it, so its checks count towards the service's availability
  • on an On-Call integration, so the alerts it raises count towards the service's response and resolution times

6. Look at the gaps

With services declared and agents reporting, open the Overview. It shows what needs attention, the applications without an owner, and recent changes. All drift opens the full drift report.

Last updated Powered by