Question 1

What are agents in SKUuz?

Accepted Answer

Agents are specialists that do catalog work for you. Each one owns a specific job — writing content, assigning categories, translating, checking quality, pricing, publishing — and runs without you driving it. Each has a trigger (automatically when products arrive, on a schedule, or when you ask) and a run mode: queue for review, where the agent holds its work for your approval, or auto-approve, where it applies directly. New agents default to review so you can watch how one behaves on your own catalog before letting it act unsupervised. Every action is recorded in the Activity Log.

Question 2

What agent types are available?

Accepted Answer

There are thirteen agent types, and every plan gives you access to all of them. Content, Translation, and Image handle enrichment. Categorization, Schema, and Validation handle structure and quality. Pricing and Publish handle commercial and channel work. Command Center, Product Operations, Import, Reporting, and Workflow handle operations — including the conversational interface that routes plain-language requests to whichever specialist should handle them. You configure when each runs and whether it needs your approval from Automate → Agents.

Question 3

Do agent runs consume credits?

Accepted Answer

Running an agent is free. Credits are consumed only when an agent uses AI — generating content, translating, classifying, or writing alt text. Rule-based work such as formula pricing and validation consumes nothing. Billing shows consumption broken down by agent, so credit usage is attributable rather than a single opaque number, and the Command Center estimates cost before starting significant work.

Question 4

What happens when I run out of credits?

Accepted Answer

Agents that require AI operations pause until credits renew at the start of your next billing cycle. Agents that don't consume credits (validation, pricing with formulaic rules) continue running normally. You can purchase a Launch Pack at any time to resume immediately, or upgrade your plan for more included monthly credits.

Question 5

Can I control whether agents auto-apply changes or require approval?

Accepted Answer

Yes. Each agent has a run mode: 'auto-approve' applies changes immediately, 'queue for review' places changes in a review queue for manual approval. You can configure this per agent, so your Content Agent might auto-approve while your Pricing Agent requires review. When queue-for-review is active, the agent's proposed changes appear in the Review Queue and on the Command Center as pending items.

Question 6

How does auto-categorization work?

Accepted Answer

The Categorization Agent assigns products to categories from your hierarchy by reading what the product already contains — its name, its vendor, how you have grouped it, and whichever additional attributes you choose to include). It gives you control over how it behaves: how cautious it should be before assigning a category, whether it may propose new categories for your approval when nothing fits, and whether its work applies directly or waits for your review. Categories are always drawn from your hierarchy — never invented — and a category you set by hand is permanent, because agents never overrule a human decision

Question 7

What is the Command Center Agent?

Accepted Answer

The Command Center Agent is your natural language interface to the entire platform. It has full context of your catalog — products, categories, attributes, agents, locales, credits, publish rules, AI content config, and BrandGuard terms. You can ask it to do anything you'd do manually: manage attributes, categories, and list values; configure publish rules and diagnose publish issues; enable and configure AI content generation; manage BrandGuard terms; re-categorize products and run confidence audits; merge categories; create and edit products; apply pricing rules; search by attributes or IDs; import files via chat upload; and more. CC understands hundreds of natural language commands across all 12 agent domains — you describe what you want, it routes to the right specialist and executes. Search actions run immediately, while data-modifying actions require your confirmation first.

Agents

Other categories

Still have questions?