> ## Documentation Index
> Fetch the complete documentation index at: https://docs-cerberus.c7-security.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Prompting guidelines

> How to steer AutoPT while a run is live - plain input versus the /pin, /unpin, and /do commands

While an AutoPT run is live, you steer it through the text field at the bottom of the session. What you type reaches the agent through one of a few **input channels**, and they differ in one important way: how long your instruction sticks around. Getting the channel right is the difference between a nudge the agent forgets on its next iteration and a fact it carries through the whole engagement.

This page covers steering a run that's already going. For setting up and launching a run, see the [Starting your first pentest](/en/starting-first-pentest) guide.

### Plain input - a one-off nudge

Type a message with no slash command and it's **ephemeral**: injected into the current iteration only. The next fresh iteration doesn't see it. AutoPT does not sort your message into anything more durable on its own - a plain prompt is always a one-off.

Use it for small, in-the-moment corrections:

* "Slow down and confirm that finding before moving on."
* "Ignore the staging subdomain for now."
* "Try a different user-agent on that endpoint."

If you want the instruction to survive into later iterations, don't type it plain - use one of the commands below.

### Making guidance stick - the slash commands

Three commands give your input a longer life than a single iteration.

#### `/pin <text>` - a standing note

A pin is re-injected into **every** iteration's prompt and shown to the agent as authoritative pinned context. This is where durable facts belong:

* credentials and tokens the agent should reuse
* leaked or hard-to-find URLs
* scope constraints ("never touch `prod-db`")
* any key fact you don't want re-discovered every iteration

```
/pin admin creds: operator / Sup3rSecret! - valid on the /admin panel
```

#### `/unpin <id>` - remove a standing note

Pins are numbered in the order you add them (1, 2, 3...). Remove one by its ID:

```
/unpin 2
```

You can also view and remove pins from the **Plugins** panel on the right (**Pins → List Pins / Remove Pin**) - the same store the `/pin` command writes to.

#### `/do <task>` - a one-shot task

`/do` hands the agent a single task, queued at high priority. It runs once, gets logged, and is not repeated. Use it to redirect attention right now without leaving a standing note behind:

```
/do focus on the admin panel and enumerate its endpoints
```

### Which channel do I want?

| You want to...                                      | Use           | Lasts                           |
| --------------------------------------------------- | ------------- | ------------------------------- |
| Nudge or correct the current step                   | plain input   | this iteration only             |
| Keep a fact in front of the agent for the whole run | `/pin`        | every iteration, until unpinned |
| Drop a standing note you no longer need             | `/unpin <id>` | -                               |
| Send a single task to run now, once                 | `/do`         | runs once, then done            |

A good habit at the start of a run: `/pin` the facts that define the engagement (credentials, scope limits, known URLs), then use plain input and `/do` to steer as you watch the thinking stream.