TotalApp Docs

AI Text Overview

Introduction to AI-powered text generation capabilities across all Writer Tools.

Overview

AI Text is the foundation layer behind all Writer Tools in TotalApp. It covers the core concepts of how the AI engine accepts prompts, generates structured content, and returns output for editing and export.

How It Works

Enter Topic
AI Generates
Review Output
Edit & Export

Each Writer Tool sends a structured prompt to the AI engine. The engine returns formatted Markdown text which is displayed in the output panel. You can edit the content inline or open it in the full Doc View editor.

Output Modes

Modal Viewer

Generated content opens in a fullscreen modal with formatting preserved. Default mode for most tools.

Terminal Inline

Content renders directly in the terminal block below the command. Toggle in Settings → Terminal → Output Display.

Doc View Editor

Opens in the Doc View split-pane editor for advanced formatting, collaboration, and export.

Frequently Asked Questions

What is the maximum output length?
Output length depends on the AI engine. The Claude API supports up to 4096 tokens per response. Multi-stage tools chain multiple requests to generate longer documents.
Can I provide style guides?
Include style instructions in your topic prompt. For example: "Write in a conversational tone, avoid jargon, use short sentences."
Which AI engines can power text generation?
Writer Tools support three engines, selectable in Settings → Writer Engine: the hosted API (Claude, default), Local CLI (uses your Claude Code subscription via the Claude CLI), and Local LLM (BERT, intent classification only — it cannot generate text). Some tools also offer a Simplified.com or Cohere sub-mode for specific writers like Blog and Storyteller.
Can I regenerate or edit the output after it's generated?
Yes. Every writer block has Edit and Regenerate actions. Editing switches the output panel to an inline editor; Regenerate re-sends the original prompt (optionally with your edited instructions) and replaces the content once the new response arrives.