Ergoki

Ergoki is a suite of tools intended to help office workers and creatives complete their tasks and build systems—going beyond the chat and “smart IDE” models, which are not adequate for office work. Simply adding an AI interface to a document suite isn’t enough; integrating AI and productivity must go further—and be more creative—in many ways.

Here are some of its guiding insights:

  • Office workers create documents: gathered, transformed, and arranged information. The AI tools that may help at each step must be differently crafted and carefully assembled in the same suite.
  • Office tasks begin by gathering relevant information. This should also be aided by AI, and often occurs before the request to accomplish a task.
  • Office workers have limited need for automated systems. Instead, they need assistance in completing tasks they personally must oversee, using information only they can identify. Such tasks are complete only after the office worker has reviewed them, based on parameters only the worker understands—often by intuition.
  • To be useful in office work, AI tools must be integrated alongside non-AI tools. For example, ordering a list alphabetically or via an AI prompt are similar tasks and should sit together. Additionally, tools of this kind must not be mere AI calls; they should verify whether the AI model skipped any items and automatically include any it missed.

Now, let’s move on to the specifics.

Let’s review the buttons on the left sidebar to understand the app’s areas and what they do.

The first icon opens the Startpage—the hub of Ergoki. Add bookmarks and access the other areas of the app.

Constructs are to Ergoki what documents are to Word or spreadsheets to Excel. A construct stores information and the tools that manipulate it—with or without AI. Each element has an order number so work progresses logically and, when needed, as a repeatable set of steps you can tweak and rerun.

Dictionaries are like spreadsheet tables, with useful additions. They organize information for your purposes. A classic example is a movie dictionary with labels such as “director,” “your rating,” “watch date,” “release year,” and “notes.” Possibilities are open‑ended—cards for a board game you’re designing, sources for a paper, daily meals for calorie tracking, and more. Dictionaries are designed to be manipulated by other tools in the app. In table terms, each row is an “entry,” and each column is a “label.”

The prompt library stores your prompts for reuse across Ergoki. You can also copy any prompt to the clipboard for use outside the system.

Systems let you process entries in a structured way. They can include AI prompts (either from the prompt library or text inside the system) and can fetch content from webpages named in entries. For example, you can create an entry with only an IMDb URL and run a system that visits the page, extracts details, sends them to AI, and fills labels such as “director” and “genre”—without relying on the model’s training data or its own internet search.

In the Files area, you’ll find all images you created in the app—either directly from a prompt or produced by a system. You’ll also find Markdown versions of information gathered elsewhere. Dictionaries and constructs can be exported as Markdown and downloaded easily. You can sync this Markdown vault with GitHub to keep it backed up. In short, Ergoki can manipulate information in your Obsidian‑style Markdown vault.

You can create AI characters and their avatars. Characters can access constructs, so you might build a dictionary, add entries, manipulate them in a construct, and then make that knowledge available to a character. Each character has parameters such as “characteristics,” “appearance,” “tasks,” “relationships,” and optional rules like “State the current date and time,” “Don’t yield under pressure,” or “Be overtly friendly.” For example, create a construct of favorite movies from a “movies” dictionary, then give a character the persona of a friend who loves the same films and chat with them—richer than a generic model.

Recollections record what you do in Ergoki: every construct and dictionary you create, each system run, your AI interactions, and your personal log. It can serve as a diary that includes daily notes alongside activity in the app itself—much as if your diary app were automatically connected to your Google Drive usage, so writing a document or sending an email could be reflected in your personal record.

This overview goes beyond the surface of what Ergoki can do, but it isn’t exhaustive. You can export entries, dictionaries, and constructs as JSON, PHP arrays, or Markdown. You can share an entire dictionary and load it with another account. Avatar sets track both emotion and location so a character’s appearance adjusts as you interact. A naming system organizes systems, dictionaries, and constructs into dynamic groups. And there are tools available in both dictionaries and constructs that don’t involve AI at all—they simply make data easier to shape and, when needed, share with AI. The rest is best discovered through use.