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

# Automate Message Rule Enforcement with Prowl AutoMod

> Prowl AutoMod scores every message against your rule set and automatically deletes, mutes, or bans based on low, medium, or high severity.

Prowl's AutoMod system watches every message sent in your server and takes action based on your configured filters — all in real time and without a moderator needing to be online. You define the filters and actions once in the dashboard, and Prowl handles the rest.

## How AutoMod Works

When a message is sent, Prowl evaluates it against each enabled filter in sequence. The first matching filter triggers an action — Prowl stops checking after the first match, so only one action fires per message.

1. **Message is received** by Prowl.
2. **Filters are evaluated** in order — profanity, spam, links, caps, mentions, invites, zalgo, emoji.
3. **First match triggers** — the matching filter's configured action is executed.
4. **Action is logged** to your AutoMod log channel with the matched filter, the message content, and the user.

<Info>
  Moderators are exempt from AutoMod by default. Members with Administrator, Manage Messages, or Moderate Members permissions — or any role in your configured mod roles — will not be filtered.
</Info>

## Filters

Prowl provides eight built-in filters, each independently configurable:

| Filter             | What It Detects                                                                                               | Default State    |
| ------------------ | ------------------------------------------------------------------------------------------------------------- | ---------------- |
| **Profanity**      | A curated list of banned words and slurs. You can add custom words or use the built-in list.                  | Enabled, delete  |
| **Spam**           | Identical or near-identical messages sent rapidly in succession (configurable: message count and time window) | Enabled, delete  |
| **Links**          | URLs not on your allowlist. Media links from Discord CDN, Tenor, Giphy, and Imgur are whitelisted by default. | Disabled, delete |
| **Excessive Caps** | Messages where more than 70% of alphabetic characters are uppercase (minimum 6 characters to trigger)         | Disabled, delete |
| **Mentions**       | Messages exceeding a configurable number of mentions (default: 5)                                             | Disabled, delete |
| **Invites**        | Discord server invite links (`discord.gg/…`)                                                                  | Disabled, delete |
| **Zalgo**          | Messages containing excessive zalgo (decorative combining) characters                                         | Disabled, delete |
| **Emoji**          | Messages containing more than the configured number of emoji (default: 10)                                    | Disabled, delete |

## Actions

Each filter can be configured with one of the following actions:

| Action          | Description                                                             |
| --------------- | ----------------------------------------------------------------------- |
| **Delete**      | Remove the offending message                                            |
| **Delete + DM** | Remove the message and send the user a DM explaining why                |
| **Warn**        | Issue a warning to the user's moderation record                         |
| **Warn + DM**   | Warn and notify the user via DM                                         |
| **Mute**        | Apply a timeout (duration configurable per filter, default: 60 minutes) |
| **Mute + DM**   | Mute and notify the user via DM                                         |
| **Kick**        | Remove the user from the server                                         |
| **Kick + DM**   | Kick and attempt to notify the user via DM                              |
| **Ban**         | Permanently ban the user                                                |
| **Ban + DM**    | Ban and attempt to notify the user via DM                               |

<Tip>
  Start with **Delete** for most filters and escalate to **Warn** or **Mute** only for repeat offenders. This keeps the experience pleasant for legitimate users who make an honest mistake.
</Tip>

## Setting Up AutoMod from the Dashboard

<Steps>
  <Step title="Navigate to AutoMod">
    Log in at [Prowl](https://prowlbot.xyz/login), choose your server, and open **AutoMod** from the sidebar.
  </Step>

  <Step title="Enable AutoMod">
    Toggle **Enable AutoMod** to the on position at the top of the page.
  </Step>

  <Step title="Configure Filters">
    For each filter type (Profanity, Spam, Links, Caps, Mentions, Invites, Zalgo, Emoji), toggle it on or off and select the action you want it to take.
  </Step>

  <Step title="Customize Filter Settings">
    Adjust thresholds for filters that support them — for example, the number of messages in the spam window, the caps percentage, the mention limit, or the emoji count.
  </Step>

  <Step title="Set a Log Channel">
    Choose an **AutoMod Log Channel** so your team can review automated actions. This can be the same as your general mod-log or a dedicated channel.
  </Step>

  <Step title="Save and Test">
    Click **Save Changes**, then test a filter in a private channel (e.g., send a message with excessive caps) to confirm the action fires correctly.
  </Step>
</Steps>

## AutoMod Logs

Every AutoMod event generates a structured log entry in your chosen log channel. Each entry includes:

* **Timestamp** of the event
* **User** who triggered the filter, with their ID
* **Channel** where the message was sent
* **Filter matched** and the action taken
* **Original message content** (truncated for very long messages)

<Tip>
  Review your AutoMod log regularly to catch false positives and fine-tune your filter thresholds. A filter that's too aggressive can frustrate legitimate members.
</Tip>