Dialog Tags

Dialog Tags are a tool for color-coding, classifying, and segmenting customer conversations in the Chats section. They allow you to organize incoming chats by business objectives, topics, funnel stages, or priority — both manually by support operators and automatically via bot constructor scenarios and AI.

1. What are Dialog Tags

Dialog tags are attached directly to a specific conversation with a customer (unlike contact tags/labels which are stored in the user's global profile).

Tag Categories and Selection Types

Tags are organized into Categories. Each category has a designated selection type:

Each tag has its own customizable name and badge color for instant visual recognition in the chat interface.

2. Where Tags are Configured

Tag and category configuration is located in each bot's settings:

1. Navigate to Bots and select your bot.
2. Open Bot Settings and click on the Dialog Tags tab.
3. In this panel you can:
   • Create categories: specify a name and choose the selection type (Single or Multiple).
   • Add tags: specify a title and choose a badge color.
   • View Tag IDs: each tag displays a #ID (e.g. #14), which is used in scenarios and AI integrations.
   • Sort: drag-and-drop categories and tags to reorder them in the menu.
   • Edit & Delete: if a tag or category is already assigned to active conversations, the system prevents accidental deletion and notifies you of existing associations.

3. How to Use Tags in Chats CRM

Assigning Tags in a Conversation

1. Open a conversation in the Chats section.
2. Click the + Tag button in the conversation header / details panel.
3. In the dropdown menu, select the desired tags. If a category is single-select, selecting a tag automatically replaces any prior tag from that category.
4. To remove a tag, click the cross icon on the tag badge or uncheck it in the dropdown menu.

Manager Activity Log: tag changes made by operators in the CRM panel do not clutter the chat transcript with system messages; instead, they are recorded in the Manager Activity audit log (tag_updated).

Display and Tooltips

• In the conversation list, tags appear as compact colored badges on each conversation card.
• If a conversation has multiple tags, they neatly collapse into a +N badge (e.g., +2). Hovering over the badge opens a tooltip with the complete list of hidden tags.

Filtering Conversations by Tags

Above the conversation list, the filter panel includes an advanced tag filter with three matching modes:

• All Tags (AND) — matches conversations that have all of the selected tags.
• Any of Tags (OR) — matches conversations that have at least one of the selected tags.
• Without Tags (Exclude) — excludes conversations containing the specified tags.

Disabling the Tags Display

If your workflow does not require tags, you can hide them from the chat interface:

• Open Chat Settings (the gear icon in the Chats header).
• Toggle the "Show dialog tags" setting ("Dialog tags are specified in the bot settings and allow you to separate dialogs by business tasks").

4. Bot Constructor: "Chat Actions" Block

In the bot constructor, the "Chat Actions" block allows you to automatically adjust the conversation state as the user progresses through a bot flow.

Inside the block, enable the "Manage Tags" section:

Adding Tags

• Select from list: pick one or more pre-configured bot tags.
• From variable: provide a variable containing an array of IDs [1, 2, 3] or a comma-separated string 1, 2, 3 (e.g. {{selected_tags}} or {{ai_tags}}).

Removing Tags

• Select from list: check specific tags to remove from the conversation.
• Clear entire category: removes all tags belonging to a selected category (useful for resetting a funnel stage).
• From variable: remove tags by IDs provided in a variable.

Unified System Message

When a bot flow executes a Chat Actions block, all triggered actions are combined into a single consolidated system message in the chat transcript, with each action printed on a new line:

Manager John joined the conversation
Conversation added to folder "Hot Leads"
Bot auto-replies disabled
Tags added: VIP, Initial Consultation
Tag removed: New Lead

5. Automated AI Tagging and Classification

You can have an LLM / AI analyze user messages or dialogue context and dynamically assign matching tags.

The workflow consists of three steps:

graph LR
    A["Database Block<br>(Get Dialog Tags List)"] --> B["AI Integration / AI Dialog<br>(Classify & Output Tag IDs)"]
    B --> C["Chat Actions Block<br>(Assign Tags via Variable)"]

Step 1. Fetch Tag List from Database

1. Add a Database block to your bot flow.
2. Select the "Tags" (or "Dialog Tags") source.
3. Select the "Get tag list" action.
4. (Optional) Select a specific tag category if you only want to classify within one group (e.g. Services).
5. Result variable name: defaults to dialog_tags (contains a JSON array of tags with their IDs and names).

Step 2. AI Analysis & Classification

In the AI Integration block (or the system prompt of an AI Dialog node), pass the tag list variable and set a strict instruction for the output.

Sample System Prompt:

You are a conversation classification assistant.
Here is the list of available tags in JSON format:
{{dialog_tags}}

Task:
Analyze the user's latest message and conversation history.
Select the matching tags from the list above.

Response format:
Return ONLY the numerical IDs of the selected tags separated by commas (e.g.: 3, 12).
If no tags match, return nothing (empty response).
Do NOT include any explanations, extra text, or punctuation other than numbers and commas.

Save the AI output to a variable, for example {{ai_tags}}.

Step 3. Assign Tags in Chat Actions

1. Add a Chat Actions block following the AI step.
2. Enable "Manage Tags" → "Add tags".
3. Set the mode to "From variable" and enter {{ai_tags}}.

The bot will automatically:

• Parse and validate the IDs from the variable.
• Assign the corresponding tags to the conversation (respecting single-selection category rules).
• Post a consolidated system message in the chat displaying the names of the assigned tags.