> For the complete documentation index, see [llms.txt](https://docs.perisclaw.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.perisclaw.com/security-and-control/permissions-and-access.md).

# Permissions & access

Perisclaw only ever does what you allow. You decide which services it connects to, which chats it can see, and what it can do in each — and you can change any of it at any time. Control stays with you, always.

## What your agent can access

Your agent only accesses what you connect and authorize, nothing more. Depending on your setup, that can include:

* **WhatsApp** — your chats, contacts, groups, media, and message history
* **Email** *(coming soon)* — messages, threads, drafts, and attachments
* **Calendar** — events, attendees, and availability
* **Files** — documents from connected storage like Google Drive and Sheets
* **Other tools** — any tool or account you explicitly connect through a connector

It uses this only to carry out what you ask — like drafting a reply, scheduling a meeting, or tracking a follow-up. See [Connecting apps](/skills-and-integrations/connecting-apps.md) to manage what's linked.

## How your agent reads and writes your chats

By default, your agent is **read-only** across all your chats. It can see them in order to do its work, but it never sends anything on its own. You can also mark any chat as sensitive or off-limits, and your agent will leave it alone entirely — neither reading nor acting in it.

When your agent does write to a chat, it happens in one of three ways:

{% stepper %}
{% step %}

### You ask it to

It drafts the message, shows it to you, and sends only once you approve.
{% endstep %}

{% step %}

### You invoke it with `/claw`

It responds directly in the chat where you called it.
{% endstep %}

{% step %}

### You set it to run autonomously

You can tell it to take over certain chats on your behalf — like every new chat that comes in today — and it handles them on its own until you stop it.
{% endstep %}
{% endstepper %}

Across all three, your agent works strictly within the permissions you've set. To learn more about these, see [Chat modes & autonomy](/security-and-control/chat-modes.md).

## It confirms before it acts

You control how much your agent does on its own. By default, anything it sends or changes on your behalf is shown to you for approval first — so a drafted reply waits until you confirm it before going out. When you want it to run a longer task without checking in at every step, you grant it that autonomy deliberately. The level is always yours to set.

{% hint style="info" %}
Even in autonomous mode, your agent never commits money, agrees to a payment, or makes a binding promise on your behalf without checking with you first. Those always come back to you for approval.
{% endhint %}

## Connect and disconnect anytime

Connect a service when you want the capability, and disconnect it the moment you don't. Disconnecting immediately stops all future access to that service. Nothing stays connected without your say-so.

## Manage your data and memory

You stay in control of everything Perisclaw holds. At any time you can:

* Delete specific data
* Clear what your agent remembers
* Delete your account entirely

To delete your account, [open Perisclaw](https://perisclaw.com), click the three dots next to your phone number, and delete it from there.

{% hint style="success" %}
Want the full picture of how your data is stored and protected? See [Privacy & security](/security-and-control/privacy-and-security.md).
{% endhint %}


---

# 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:

```
GET https://docs.perisclaw.com/security-and-control/permissions-and-access.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
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.