Documentation Index

Fetch the complete documentation index at: https://docs.unqork.io/llms.txt

Use this file to discover all available pages before exploring further.

UnqorkAI Overview

Prev Next

UnqorkAI brings AI-native capabilities to the Unqork platform at two levels. The Build Agent helps Creators build applications. AI components are embedded in applications to give end-users AI-powered experiences.

UnqorkAI Build Agent

The UnqorkAI Build Agent is integrated directly into the Module Builder. Creators describe what they want to build in plain language. The agent creates a plan, presents it for approval, then builds the components, connects the logic, and runs tests automatically.

The Build Agent operates inside Unqork's established guardrails, using the platform's proven components and configuration patterns. The result is predictable, governed, and auditable output with no freeform code.

The Build Agent can do the following:

  • Generate forms, input flows, and data displays from a plain-language description.

  • Implement calculations, filtering, and data transformations through natural language prompts.

  • Connect components by wiring triggers, watchers, and data flows.

  • Debug module configurations and explain what components do.

  • Generate test cases and verify logic in API modules.

  • Read other modules in the application to inform what it builds in the current module.

New to the Build Agent?

AI tools in a no-code environment are new territory for most Creators. The following articles provide a guided introduction, best read in order:

  1. Getting Started with the Build Agent — opens the agent, sends a first prompt, and walks through the plan-approve-review cycle.

  2. Using the Build Agent Effectively — how to prompt well, what to verify, what the agent can and cannot see, and how to keep sessions productive.

  3. Implementing Logic with the Build Agent — once the basics are comfortable, this covers data processing and logic implementation in depth.

The Build Agent Reference is a complete reference for all features. It is not required reading to get started. Creators can return to it when looking up a specific capability.

Article

Description

Getting Started with the Build Agent

First session walkthrough — opening the agent, sending a prompt, and reviewing the result.

Using the Build Agent Effectively

Prompting strategies, session management, verification, hallucinations, and privacy.

Implementing Logic with the Build Agent

Calculations, transformations, and data processing through natural language.

Build Agent Reference

Full reference for all agent capabilities, Tasks, Knowledge Nuggets, Plans, and data privacy.

AI Components

AI components are components Creators add to their modules to give end-users AI-powered experiences. Each AI component runs on an LLM credential the organization registers. Administrators configure credentials under Administration > Environment Settings > LLM Settings > LLM Credentials.

Component

Description

AI Summarizer Component

Reads and summarizes content in the application. Condenses documents, submissions, or data sets into structured output for end-users or Creators.

For guidance on configuring model-level settings like Temperature, Thinking Level, and Max  Tokens, see Understanding AI Model Settings.

Build Agent vs. AI Components

The Build Agent and AI components serve different purposes and different audiences. The table below outlines the key distinctions.

Build Agent

AI Components

Who uses it

Creators building applications

End-users using applications

Where it lives

Module Builder, Build Agent panel

Inside modules, visible to end-users

What it does

Builds and configures the application

Provides AI-powered experiences in the application

Context it sees

Module configuration and structure

Module submission data and component context


Changelog

Date

Change

2026-08-19

Updated AI Components section: LLM Credential now required; removed Gemini-managed reference.

2026-07-01

Initial publication.