> For the complete documentation index, see [llms.txt](https://docs.lassomoderation.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.lassomoderation.com/getting-started/key-concepts.md).

# Key Concepts

Lasso is built around a small set of concepts that come back everywhere: in the dashboard, in the API and in webhooks. This page introduces each concept and links to the guide that covers it in depth. If you are new to Lasso, start here, then read the [Lasso Moderation Overview](/getting-started/lasso-moderation-overview.md) to see how content flows through the platform in practice.

### Workspace

Your platform's environment in Lasso. A workspace holds your content, users, rules, policies, review queues, team and settings. API keys and your webhook are configured per workspace under **Settings → API & Webhooks**.

### Categories and subcategories

Two levels of structure that mirror how your own platform is organized:

* **Category**: a top-level area of your platform, for example a community, a game or a marketplace section.
* **Subcategory**: a place within a category where content is posted, for example a discussion thread, a chat room or a listing.

Every piece of content belongs to a subcategory, and every subcategory belongs to a category. You can use them to filter content, build rule conditions and limit which parts of your platform each moderator can access.

### Content

A single item created by one of your users: a message, a post, a listing or an image. One content item can combine text with images, videos and audio files. Content enters Lasso through the [API](https://api-reference.lassomoderation.com/) or through an integration, and every item is analyzed and checked against your rules as it arrives.

{% hint style="info" %}
Text and image moderation are available in every workspace. Video and audio moderation are enabled per workspace, contact us to turn them on. See [Image Moderation](/content-analysis/image-moderation.md) and [Video Moderation](/content-analysis/video-moderation.md).
{% endhint %}

### Users

The people on your platform. Lasso builds a profile for each user from the information you send: their content, signup date, country, tags, custom attributes and more. Users are moderated just like content, manually or automatically, and can also receive temporary bans and strikes. See [User Moderation](/managing-users/user-moderation.md).

### Statuses

Every content item and every user has one of three statuses:

* **Allowed**: fine to stay on your platform, no action needed.
* **Flagged**: needs a closer look. Flagging an item creates a review for your moderators.
* **Removed** (for users: **Banned**): should be taken down or blocked on your platform. In the API this status is called `hidden`.

Lasso records both the initial status (the outcome of the automated checks when the item arrived) and the current status (after any later decisions), so you can always compare automated and human outcomes. Status labels are configurable per workspace, so your dashboard may show other words, such as "Hidden", "Suspended" or "Deleted".

### Actions

Every moderation decision is recorded as an action: who took it (a moderator, a rule or an AI Moderator), what changed and why. Actions appear on the item's activity timeline and in the audit logs, and they are delivered to your platform through [webhooks](/developers/webhooks.md) so your platform can enforce them. Lasso never changes anything on your platform directly; your platform decides what a removal or a ban means.

### Policy categories and policies

The **Policies** section of the dashboard contains two related concepts:

* **Policy categories**: the reasons attached to decisions. When a moderator, rule or AI Moderator takes an action, the policy category records which guideline was violated, for example "Hate speech" or "Spam". A category's identifier appears as `policy_id` in the API and in webhooks.
* **Policies**: written policy documents describing your platform's guidelines. AI Moderators read a policy document to decide how to moderate. Policies are versioned, so you can draft changes and publish them when ready.

### Rules

Automated moderation you control. A rule combines conditions with actions: when all conditions match, the actions run. Rules evaluate on incoming content and users, and can change a status, attach a policy category, add tags, give strikes or route reviews to a specific queue. They can be adjusted on the fly, without code changes or deployments. See [Custom Moderation Rules](/automation-and-review/custom-moderation-rules.md).

### Lists

Reusable sets of words, domains, countries, phone numbers, IP addresses or regex patterns that rule conditions match against. Update a list once and every rule that uses it follows. See [Lists](/automation-and-review/lists.md).

### AI analysis and classifiers

Lasso analyzes every incoming item automatically: text gets AI scores for toxicity and related categories, images are scanned for unsafe content and for text they contain, and classifiers check content against policies specific to your platform, answering a single question: does this item violate the policy or not? These signals never act on their own. Rules turn them into decisions, so you stay in control of the thresholds and the consequences.

### AI Moderators

AI members of your moderation team. An AI Moderator reads your written policy and judges content against it the way a human moderator would. Depending on its configuration it takes action directly, with the policy category and its reasoning attached, or leaves a recommendation for a human to confirm. You control which actions it may take and how decisive it is (**Cautious**, **Balanced** or **Decisive**) on the **Team** page.

### Reviews and review queues

A review is one item waiting for a human decision. Flagged content and users become reviews, and review queues organize those reviews for your moderators. Routing rules decide which queue each review lands in; anything not matched by a routing rule goes to the Default Queue. Reviews are assigned automatically to the moderators working in a queue. See [Review Queues](/automation-and-review/review-queues.md) and [Routing Rules](/automation-and-review/review-queues/routing-rules.md).

### User reports

Reports filed by your users about content or other users, sent to Lasso through the API. Reports show up in **User Reports** in the dashboard, and each report also re-evaluates your rules, so a rule can act on heavily reported items within seconds. See [User Reports](/automation-and-review/user-reports.md).

### How a decision is made

When a new piece of content arrives, Lasso evaluates it in a fixed order:

1. **Rules run first.** A matching rule can allow, flag or remove the item, attach a policy category, add tags or give a strike.
2. **The AI Moderator runs second.** When it is configured to check incoming content and no rule has already flagged or removed the item, it takes action or leaves a recommendation.
3. **Nothing matched?** The item is allowed.
4. **Flagged items become reviews** and are routed to a review queue for a human decision.
5. **Every action is sent to your platform** through your webhook.

### Tags, custom attributes, strikes and messages

Tools that support day-to-day moderation:

* **Tags**: your own labels on users and content, applied manually or by rules, and usable as filters and rule conditions. See [Tags](/moderator-toolkit/tags.md).
* **Custom attributes**: extra data fields you send along with content and users, such as "Verified" or "Age", shown in the dashboard and usable in rule conditions. See [Custom Attributes](/moderator-toolkit/custom-attributes.md).
* **Strikes**: escalating temporary penalties for repeat offences, instead of an immediate permanent ban. See [Strikes](/managing-users/strikes.md).
* **Messages**: messages sent to your users from Lasso, to explain decisions or share guidelines. See [Sending Messages](/moderator-toolkit/sending-messages.md).

### The API and webhooks

Content and users flow into Lasso through the REST API, and decisions flow back through your webhook:

* `POST /api/v1/content` sends content asynchronously; the decision arrives later through your webhook. This is the standard setup for post-moderation.
* `POST /api/v1/content/sync` waits for the decision and returns it in the response, for pre-moderation setups where content only goes live after approval.
* Requests authenticate with an API key, created under **Settings → API & Webhooks**.
* The webhook delivers every action with a signature you can verify, so your platform can safely enforce decisions. See [Webhooks](/developers/webhooks.md).

The full API is documented in the [API Reference](https://api-reference.lassomoderation.com/).

### Team and roles

The **Team** page is where you invite moderators, configure AI Moderators and control access. Admins manage settings and the team, moderators handle day-to-day moderation, and a moderator can be restricted to specific categories when they should only see part of your platform.