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

# Agents inventory

> Read the Agents inventory: what each summary tile counts, what each column and filter means, and how search, sorting, and empty values work.

The Agents inventory is one list of the AI agents Oleria discovered across your connected applications. Use it to see which agents exist, narrow the list to the ones that need attention, and open any agent to see its accounts, its access, and who can use it.

To open it, go to **Agents** -> **Inventory**. In the classic navigation, open **Access Inventory** and select the **Agents** tab.

> [!NOTE]
>
> Agent discovery is enabled per workspace. To turn it on, contact us at [support@oleria.com](mailto:support@oleria.com).

The inventory lists agents in every lifecycle state, including suspended and archived ones. Select an agent's name to open its [details page](https://docs.oleria.com/governance/agent-details.md).

## Summary tiles

The **Summary** section shows five counts. Each tile counts only the agents that match the filters applied to the list, so the numbers change as you filter.

| Tile | What it counts | Selecting it |
| :- | :- | :- |
| **Total Agents** | Every agent that matches the current filters. | Can't be selected. |
| **Critical or High Risk** | Agents whose risk level is **Critical** or **High**. | Filters the list to **Risk** of **Critical** or **High**. |
| **No Accountable Owner** | Agents with no [accountable owner](https://docs.oleria.com/governance/agents-overview.md#key-terms). | Filters the list to **Has Accountable Owner** of **False**. |
| **Externally Exposed** | Agents anyone can use without signing in, or that someone outside your organization can use. | Filters the list to **Externally Exposed** of **True**. |
| **Shared Account** | Agents that run as an account at least one other agent also runs as. | Filters the list to **Shared Account** of **True**. |

If the list is already filtered on a tile's field with a different value, the tile can't be selected until you clear that filter. Its help text says which one, for example **Clear the Risk filter to open this card.**

## Columns

The table shows seven columns by default. Use **Manage columns** to show, hide, or reorder columns. **Reset to default** restores the default set. The **Agent** column can't be hidden.

| Column | Default | What it shows |
| :- | :- | :- |
| **Agent** | Shown | The agent's display name, or its source name when it has no display name. When the two differ, the source name appears beneath the display name. |
| **Application** | Shown | The application the agent was discovered in. |
| **Type** | Shown | The source application's own name for the kind of agent record, for example **Agent Runtime** or **Agentforce Service Agent**. See [Where agents come from](https://docs.oleria.com/governance/agents-overview.md#where-agents-come-from). |
| **Lifecycle State** | Shown | The operating state the source application declares for the agent. See [Lifecycle state](#lifecycle-state). |
| **Risk** | Shown | The agent's risk level. Point to the tag to list its risk reasons. See [Agent risk](https://docs.oleria.com/governance/agent-risk.md). |
| **Owners** | Shown | The agent's owners, leaving out the ones Oleria has found aren't accountable. Up to two names appear, and **+N more** lists the rest. |
| **Agent Identities** | Shown | The accounts the agent runs as. Up to three names appear, and **+N more** lists the rest. An account with no name shows as **Unnamed account**. |
| **Application Instance** | Hidden | The connected instance of the application the agent was discovered in. |
| **Approval State** | Hidden | Where the agent stands in the source application's own approval workflow. See [Approval state](#approval-state). |
| **Models** | Hidden | The models the agent and its live versions use, leaving out disabled ones. |
| **Audience** | Hidden | The widest audience the source application says may interact with the agent. See [Audience](#audience). |
| **Autonomy Level** | Hidden | How independently the source application says the agent acts. See [Autonomy level](#autonomy-level). |
| **Controls** | Hidden | The guardrails and observers that apply to the agent and its live versions, leaving out disabled ones. |
| **Hosting Model** | Hidden | Where the agent runs. See [Hosting model](#hosting-model). |
| **Protocols** | Hidden | The protocols over which the agent accepts calls. See [Protocols](#protocols). |
| **Created On** | Hidden | The date the agent was created in the source application. |
| **Modified On** | Hidden | The date the agent was last modified in the source application. |
| **Live Versions** | Hidden | The versions of the agent that are currently live. |
| **Created By** | Hidden | Who created the agent, as the source application records it. When Oleria doesn't hold that account, the source application's identifier for the creator appears. |
| **Last Activity** | Hidden | The latest activity on the agent's own accounts. [Key terms](https://docs.oleria.com/governance/agents-overview.md#key-terms) defines which accounts count. |

> [!NOTE]
>
> The **Owners** column leaves out every owner Oleria has found isn't an [accountable owner](https://docs.oleria.com/governance/agents-overview.md#key-terms). An account that only created the agent is left out this way, and appears under **Created By**. An agent whose owners were all left out shows **None**.
>
> An owner Oleria hasn't evaluated yet stays listed. While the agent's **Risk** shows a dash, that is every owner, including an account that only created the agent. After that, it is a group whose members Oleria hasn't counted yet. See [Owners](https://docs.oleria.com/governance/agent-details.md#owners).

## How to read empty values

| On screen | Meaning |
| :- | :- |
| A dash | Oleria has no value. Either the integration doesn't read this field from the source application, or Oleria hasn't computed it yet. Owners, risk, models, controls, live versions, and last activity are computed when the agent's integration syncs. In **Agent Identities**, a dash means the agent has no account of its own. In **Last Activity**, a dash also appears when every account the agent runs as is shared with another agent or can be acted as by another principal. |
| **None** | Oleria looked and found nothing: no accountable owner, no model, no control, no live version, or an agent that accepts no calls. |
| **Unavailable** | Oleria read the agent, and the source application either doesn't state this field or states a value Oleria doesn't recognize. |

## Filters

No filter is applied when you open the inventory. Filters and the column selection are kept in the page address, so a link you copy opens the same view.

| Filter | Matches |
| :- | :- |
| **Agent** | Text that the agent's source name contains or doesn't contain. |
| **Agent Identities** | Agents that run as any of the accounts you select. |
| **Application** | Agents discovered in the applications you select. |
| **Application Instance** | Agents discovered in the connected instances you select. |
| **Approval State** | **Approved**, **Draft**, **Rejected**, or **Submitted**. |
| **Audience** | **External**, **Internal**, or **Public**. |
| **Autonomy Level** | **Autonomous**, **Semi-Autonomous**, or **Supervised**. |
| **Controls** | Agents that use any of the guardrails or observers you select. |
| **Created On** | A creation date between two dates, before or after a date, or in the last **1 day**, **7 days**, **14 days**, **30 days**, **120 days**, or **1 year**. |
| **Display Name** | Text that the agent's display name contains or doesn't contain. |
| **Externally Exposed** | **True** for agents anyone can use without signing in, or that someone outside your organization can use. **False** for agents that aren't. |
| **Has Accountable Owner** | **True** for agents with at least one [accountable owner](https://docs.oleria.com/governance/agents-overview.md#key-terms). **False** for agents with none. |
| **Hosting Model** | **Embedded**, **Local**, **Platform Hosted**, or **Self Hosted**. |
| **Last Activity** | The agent's last activity date, with the same choices as **Created On**. |
| **Lifecycle State** | **Active**, **Archived**, **Established**, **Quarantined**, or **Suspended**. |
| **Models** | Agents that use any of the models you select. |
| **Modified On** | The last modified date, with the same choices as **Created On**. |
| **Orchestration Role** | **Coordinator**, **Router**, **Standalone**, or **Worker**. See [Orchestration role](#orchestration-role). |
| **Owners** | Agents that have any of the owners you select. |
| **Protocols** | Agents that accept calls over any of **A2A** (Agent2Agent), **AG-UI** (Agent User Interaction), **HTTP**, or **MCP** (Model Context Protocol). See [Protocols](#protocols). |
| **Risk** | **Critical**, **High**, **Moderate**, or **Low**. |
| **Risk Finding** | Agents that have any of the risk reasons you select. See [Agent risk](https://docs.oleria.com/governance/agent-risk.md). |
| **Shared Account** | **True** for agents that run as an account another agent also runs as. **False** for agents that don't. |
| **Type** | Agents of the types you select. |

Each filter offers the operators that fit its values:

| Operator | Filters | What it does |
| :- | :- | :- |
| **includes** or **excludes** | Every filter where you select values from a list, except **Risk Finding** | **includes** matches agents that have any of the values you select. **excludes** leaves out agents that have any of them. |
| **includes** | **Risk Finding** | Matches agents that have any of the risk reasons you select. This filter has no **excludes**. |
| **is** | **Externally Exposed**, **Has Accountable Owner**, and **Shared Account** | Matches **True** or **False**. |
| **contains** or **does not contain** | **Agent** and **Display Name** | Matches the text you type. |
| **between**, **in the last**, **before**, or **after** | **Created On**, **Last Activity**, and **Modified On** | Matches dates, as the **Created On** row describes. |

Three behaviors to know when you filter:

* The **Owners** filter matches every owner, including the agent's creator and owners who don't count as accountable. It can return an agent whose **Owners** column reads **None**.
* **Unavailable** is never offered as a filter option.
* The **Risk** filter has no option for **No known risk**.

## Search and sort

The **Search agent name** box matches the agent's source name, the same as the **Agent** filter. The **Agent** column shows the display name first, so a search can miss an agent whose display name differs from its source name. To search display names, use the **Display Name** filter. The classic navigation has no search box. Use the **Agent** filter there.

The list opens sorted by **Risk**, highest first: **Critical**, **High**, **Moderate**, **Low**, then agents with **No known risk**. Select a column heading to sort by it. Sorting by **Agent** orders the list by source name, the name that search matches, so display names can appear out of alphabetical order. You can sort by every column except **Application Instance**, **Owners**, **Agent Identities**, **Models**, **Controls**, **Protocols**, **Live Versions**, and **Created By**.

The list shows 15 agents per page. An empty list reads **No Data**.

## Field values

These fields record what the source application declares about an agent. They are the source application's statements, not Oleria's assessment. The same values appear on the [agent details page](https://docs.oleria.com/governance/agent-details.md).

### Lifecycle state

| Value | Meaning |
| :- | :- |
| **Active** | In service and able to act. |
| **Established** | Registered but not yet activated. |
| **Suspended** | Stopped, and able to return to service. |
| **Quarantined** | Frozen pending a security review. |
| **Archived** | Authority permanently withdrawn. The agent can't return to service. |
| **Unavailable** | The source application exposes no lifecycle state. |

### Approval state

Approval state records a decision made in the source application's own approval workflow, not whether the agent is able to act. The lifecycle state records that.

| Value | Meaning |
| :- | :- |
| **Approved** | The source application records the agent as approved for use. |
| **Draft** | The agent was created but not submitted for approval. |
| **Submitted** | Approval was requested and not yet decided. |
| **Rejected** | The source application records a decision against the agent. |
| **Unavailable** | The source application has no approval workflow, or has one whose state can't be read. |

### Audience

Each audience includes the narrower ones.

| Value | Meaning |
| :- | :- |
| **Internal** | Only people in your organization may interact with the agent. |
| **External** | Identified people outside your organization, such as customers, partners, and guests, may also interact with it. |
| **Public** | Anyone may interact with it without identifying themselves. |
| **Unavailable** | The source application doesn't declare an audience. |

### Autonomy level

The inventory labels this field **Autonomy Level**. The agent details page labels it **Autonomy**.

| Value | Meaning |
| :- | :- |
| **Autonomous** | The agent acts without a person involved in each action. |
| **Semi-Autonomous** | The agent acts alone for some actions and needs approval for others. |
| **Supervised** | Every consequential action needs a person's decision. |
| **Unavailable** | The source application doesn't declare an autonomy level. |

### Orchestration role

The agent's position among other agents. It appears as a filter here and as a field on the agent details page.

| Value | Meaning |
| :- | :- |
| **Coordinator** | Delegates work to other agents and also acts itself. |
| **Router** | Only delegates work to other agents. |
| **Worker** | Another agent delegates work to it. |
| **Standalone** | Neither delegates nor is delegated to. |
| **Unavailable** | The source application doesn't declare a role. |

### Hosting model

| Value | Meaning |
| :- | :- |
| **Embedded** | The agent is a feature of another vendor's application. |
| **Local** | The agent runs on an end-user device. |
| **Platform Hosted** | The agent runs on a vendor's agent platform that you configure. |
| **Self Hosted** | The agent runs on infrastructure you operate. |
| **Unavailable** | The source application doesn't declare where the agent runs. |

### Protocols

| Value | Meaning |
| :- | :- |
| **A2A** | The agent accepts calls over the Agent2Agent (A2A) protocol. |
| **AG-UI** | The agent accepts calls over the Agent User Interaction (AG-UI) protocol. |
| **HTTP** | The agent has a plain HTTP or REST interface. |
| **MCP** | The agent is itself exposed as a Model Context Protocol (MCP) server. |
| **Unavailable** | The agent can be called, over a protocol the source application doesn't name or that isn't in this list. |
| **None** | The source application says the agent accepts no calls. |

## Contact us

For questions, contact us at [support@oleria.com](mailto:support@oleria.com).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.