> For the complete documentation index, see [llms.txt](https://docs.nexos.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.nexos.ai/workspace/agents/human-in-the-loop-hitl-approvals.md).

# Human-in-the-Loop (HITL) Approvals

Human-in-the-Loop (HITL) Approvals lets your agent pause before taking an action that changes something in another tool. Instead of acting immediately, nexos.ai shows you what the agent plans to do and waits for your decision.

This helps you stay in control of actions like sending a message, creating a ticket, or updating a record.

<figure><img src="/files/GAmYZjMAYSBA2LNW1gTQ" alt="" width="563"><figcaption></figcaption></figure>

### Where to find it

You will see approvals in different places depending on how the agent is running.

#### For chat agents

The approval appears directly in the conversation when the agent reaches an action that needs your review.

#### For scheduled agents

The approval appears in the agent’s Activity history. You may also notice it from the Activity panel, where paused agents show that they are waiting for approval.

To get there:

1. Open Agents.
2. Find the agent you want to review.
3. Open the agent’s Activity history or select the notification from the Activity panel.
4. Review the pending approval card.

### How it works

When an agent is about to perform an action that changes something outside nexos.ai, it pauses and shows an approval card.

The card gives you a clear summary of:

* what the agent wants to do
* which integration it will use
* the key details of the action

You can then choose what should happen next.

### Turn approvals on or off for each integration

Approval settings are managed per integration inside the agent settings. For each integration added to the agent, you can turn the approval toggle on or off.

If the toggle is on, nexos.ai will ask for approval before the agent takes an action that changes something in that integration.\
If the toggle is off, the agent can continue with write actions in that integration without asking each time.

To manage it:

1. Open Agents.
2. Select the agent you want to update.
3. Open Studio in the top-right corner.
4. Select Settings.
5. Go to Integrations.
6. For each added integration, turn Require my approval before taking each action on or off.

<figure><img src="/files/3Qr84WHrl8cQeZryHAxN" alt="" width="563"><figcaption></figcaption></figure>

### What you can do

Each approval card gives you three options:

* Approve\
  The agent performs the action and continues.
* Decline\
  The agent does not perform the action and stops that run.
* Always approve\
  Future actions for that agent + integration can continue automatically without asking each time.

### What happens next

After you make a decision:

* if you approve, the action is completed and the agent continues
* if you decline, the action is canceled and the run stops
* if you choose always approve, the action is completed and similar future actions for that agent/integration can continue automatically

If no action is taken, the request will eventually expire and the run will stop.

### What users need to know

* Approvals only appear for actions that change something in another system.
* Read-only actions, like searching or retrieving information, continue without asking for approval.
* Approval requests stay available even if you leave and come back later.
* You should always review the action before approving it.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.nexos.ai/workspace/agents/human-in-the-loop-hitl-approvals.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.