Skip to main content
SketricGen has two product paths, but the same core concepts power both.

​
Brand Agents

Brand Agents are customer-facing agents trained on your brand, website, and approved knowledge. They are built for public deployment through widgets or customer channels. They are the best fit for a website chatbot, lead capture assistant, or visitor-facing support agent. Brand Agents focus on security, brand behavior, lead capture, visitor guidance, and controlled tool access.

​
AI Workforce

AI Workforce is the internal workspace for multi-agent work and workflow automation. It uses Agent Spaces and the updated runtime for longer runs, planning, files, artifacts, tools, and team workflows. AI Workforce is the right place for research, content, operations, sales, internal support, and file-heavy work.

​
Agents

An agent is a role with:
  • Instructions
  • A selected model
  • Tools it can use
  • Optional structured output
  • A place in the workflow or product surface
Good agents have narrow responsibilities and clear success criteria.

​
Tools and Connectors

Tools let agents act or retrieve information. Examples include:
  • Knowledge Bases
  • Web Search
  • Image Generation when enabled
  • API Request
  • 2,000+ SketricGen connectors
  • Marketplace skills
  • Custom MCP servers
  • Brand Agent Lead Capture and Page Navigation
Only assign tools that the agent truly needs.

​
Skills

Skills are reusable task packages for AI Workforce agents. They can package task-specific instructions, files, and operating patterns so an agent can do a job in a more reliable way than starting from a blank prompt. Use skills when you want an agent to follow a curated process, such as analyzing a document, creating a report, reviewing a design, planning a project, or producing an artifact.

​
Knowledge Bases

Knowledge Bases store trusted information from websites, files, text, or Q&A pairs. Brand Agents use them for public answers. AI Workforce uses them for internal reference data.

​
Main Agents and Sub Agents

In AI Workforce, the main agent owns the user conversation and can call specialist sub agents for bounded work. Use sub agents for classification, extraction, summarization, validation, research, and other focused helper tasks. Sub agents cannot have their own sub agents, so workflows stay one level deep and easier to debug.

​
Runs, Schedules, Files, and Artifacts

AI Workforce runs can create and inspect files. A run workspace stores conversation-specific artifacts, while an agent workspace can store reusable files for future runs. Schedules can trigger AI Workforce runs automatically for recurring internal work, such as daily briefs, weekly reports, or periodic checks across connected tools.

​
Traces

Traces show how a run executed: agents, sub-agent calls, tool calls, errors, latency, credit usage, and generated artifacts. Use traces to debug and improve agents before relying on them.