Resources

The glossary explains the platform's key terms in simple words.

Glossary

• Agent – an AI program specialized in a specific task – like an employee who is responsible for exactly one area of work. For this, it receives a fixed instruction, matching knowledge, and capabilities (see Introduction to incoreon AI).
• Agent library – collection of publicly available agent templates: Every template contains a finished configuration of instruction and capabilities and is adopted as an own copy (see Using the Agent Library and Templates).
• Agent storage – an agent's structured data store: Information is stored as entries with fixed fields – like rows in a table – and remains available across conversations (see Agent Storage and Data Integration).
• Agents overview – overview of all assigned agents as cards with name and description; management rights additionally show buttons for creating and for the agent library (see Understanding Agents).
• Instruction – the permanently stored base description of an agent: Who is it, what should it do, in what frame? It applies to every conversation (see Agent Components in Detail).
• Trigger – event that starts an automation: an incoming e-mail of a mailbox (“Inbox”) or a due schedule (“Schedule”) (see Automation and Workflows).
• Automation – rule-based workflow that executes an agent without a manual request, e.g. on a schedule (see Automation and Workflows).
• Automation chat – chat that every execution of an automation creates; it shows what the agent did on the run (see Automation and Workflows).
• Builder license – license tier that allows creating and managing agents (see Roles and Permissions).
• Chat – central interface for exchanging with agents (see Chat).
• Three-dot menu – menu with further actions next to an entry, e.g. a chat or a to-do; opened via the symbol with three dots (see Working on ToDos).
• Decision to-do – to-do type that requires a decision between two options; it is not checked off, but completed via the decision question (see Understanding To-Do Types).
• Capability – function an agent can use, e.g. email or browser (see Agent Components in Detail).
• Error ID – unique identifier of an error message; the basis of every support request (see Error Messages and Their Meaning).
• Thought process – the steps of thought the agent goes through before its answer; they are shown in the answer in a dashed speech bubble. Technically also called reasoning (see Chat).
• Shared account – login mode of a capability: All users of an agent use the same access with the service, e.g. a shared mailbox (see Choosing the Login Mode: Shared or Personal Account).
• Context limit – maximum amount of text an agent can process in a chat; also called context window. The context window is the agent's short-term memory – everything that has been said or uploaded in the ongoing conversation. Once it is full, a new conversation with a short summary of the most important results is recommended.
• Information – short message or update from agents, from automations, or from colleagues; it accumulates centrally in the “Information” area (see Information).
• Internet search – research on the internet without an agent: With internet search enabled, the chat searches the internet instead of agent knowledge and answers with source references (see Selecting an Agent or Using Internet Search).
• Cost limit – an agent's monthly budget; once reached, the agent stops working (see Cost Control).
• Lite – lowest of the three license tiers of incoreon AI (“Lite”, “Pro”, “Builder”); the license determines which functions a person can use (see Roles and Permissions).
• MCP – abbreviation for Model Context Protocol: the standardized connector through which incoreon connects additional functions and external systems – comparable to a USB port (see Agent Components in Detail).
• MCP server – service that provides a capability for incoreon via MCP; the capability includes various tools (see Agent Components in Detail).
• Multi-agent system – association of several specialized agents that collaborate. The central form in incoreon is the sub-agent system (see Multi-Agent Systems).
• Network drive – shared folder in the network whose files incoreon regularly synchronizes as an agent's knowledge (see Agent Components in Detail).
• Orchestrator chat – chat in which an agent controls an automated process; marked in the chat history and reachable from a to-do via “To orchestrator chat” (see Viewing the Orchestrator Chat).
• Personal account – login mode of a capability: Every person connects their own account via the profile; the agent works under their identity (see Profile).
• Pro – middle license tier of incoreon AI; it includes, among others, the chat history (see Roles and Permissions).
• Project – folder in the chat history that bundles chats on one topic or customer (see Organizing Chats in Projects).
• Prompt / prompting – the text input in the chat: a question or task for the agent. Skillfully phrasing this input is called prompting (see Chat).
• Process to-do – to-do type whose processing returns a result to the agent: It is processed in an own chat; when it is completed, the chat history goes back to the agent and the workflow continues (see Understanding To-Do Types).
• Language model – the software inside an agent. It understands text input and generates answers (see Agent Components in Detail).
• Sub-agent – agent that is integrated by another agent for subtasks (see Multi-Agent Systems).
• Sub-agent system – a controlling agent with the sub-agents integrated with it (see Multi-Agent Systems).
• System capability – always-active capability of every agent; it provides the basic functions of the collaboration, e.g. user search and creating to-dos and information (see The System Capability).
• Token – unit of measurement for a language model's consumption; the basis of cost calculation (see Cost Control).
• To-do chat – chat in which a to-do is processed; marked in color in the chat history (see Recognizing Chat Types).
• ToDos – tasks with deadline and responsible person, created by agents, from automations, or by colleagues; tasks due today are highlighted in red (see ToDos).
• Tools – individual functions within a capability integrated via an MCP server. The MCP server provides the capability; its tools can be enabled or disabled individually (see Guide: Adding a Custom MCP Server).
• Workflow orchestrator – the controlling role of the agent in an automated process: The agent creates to-dos and information, collects results, and derives follow-up steps (see How a Workflow Runs).
• Knowledge – files and network drives that are available to an agent in addition (see Agent Components in Detail).
• Knowledge group – named group of files in an agent's knowledge; its description helps the agent find the right knowledge for a matching question (see Adding Knowledge).