> ## Documentation Index
> Fetch the complete documentation index at: https://docs.endorlabs.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Agent activity

> <Badge color="green">Beta</Badge> <br /> Read the usage metrics the Agents Hub reports for each agent in your tenant.

The Agents Hub shows you how your team uses agents in your tenant, so you can see which agents are active, how often they run, and when they last ran. This page explains what those numbers mean and how to read them.

<Warning>
  Agent activity is usage metering, not a compliance audit log. The data is approximate, and Endor Labs refreshes it periodically, so it can lag real activity. Don't treat it as your system of record for audit or compliance.
</Warning>

## View agent activity

Activity appears on the agent cards in the catalog itself, not on a separate page.

1. Select **Agents Hub** from the left sidebar.
2. Look at the **Active agents** section at the top of the catalog. Each card there shows the agent's activity.

If you don't see an **Active agents** section, no agent in your tenant has run recently, and the catalog shows the full agent list instead.

## What the activity view shows

Each agent that ran recently appears under **Active agents** with two indicators on its card.

* **Tasks**: How many calls the agent made to the Endor Labs API.
* **Last run**: When the agent most recently made a call.

An agent counts as active when it made at least one call to the Endor Labs API in the reporting window. An agent with no recent activity appears under **More agents** without these indicators.

The view covers the last 7 days of activity, and Endor Labs retains activity for up to 30 days.

The view counts only the requests agents make with delegated credentials. Requests you run yourself, such as plain `endorctl api` commands, don't appear here. See [Delegated agent credentials](/secure-ai-coding/agents-hub/delegated-credentials) to learn how agents authenticate.

## How this differs from coding agent governance

The Agents Hub activity view and coding agent governance answer two different questions.

* **Agent activity** in the Agents Hub is what an agent called against the Endor Labs API, seen from the backend.
* **Coding agent governance** is what an agent did on the developer's machine, such as the tools it invoked, the commands it ran, and any policy violations during a coding session.

Use agent activity to understand API usage per agent. Use coding agent governance to understand on-device behavior.

## Data isolation

Endor Labs scopes activity to your tenant. You see usage for your own tenant only, never another tenant's agents or calls. The view aggregates activity across all namespaces in your tenant, and you can't narrow it to a single namespace.