Sage Openui logo

Sage Openui

CommunityPopular
ZHangZHengEric
sage-openui

Present answers as streaming native interactive UI or presentations in Sage Desktop v2. Use when structured visual presentation or user interaction helps accomplish the task.

Overview

PublisherZHangZHengEric
RepositorySage
Skill namesage-openui
Stars
1.2K
Forks
113
Bundled files
1
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 1 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by ZHangZHengEric on GitHub. Read the source before you install it.

Installation

Install the Sage Openui AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/ZHangZHengEric/Sage/tree/main/app/skills/sage-openui
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/ZHangZHengEric/Sage.git /tmp/Sage
mkdir -p .claude/skills
cp -r /tmp/Sage/app/skills/sage-openui .claude/skills/sage-openui
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Sage Openui in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Sage Openui on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Sage Openui is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Sage OpenUI

Capability boundary

Interactive UI is a built-in answer format. Generate it directly in an assistant message inside a fenced code block whose language is exactly openui. The host parses incoming message chunks with the official parser and renders native Flutter Glass components.

Rendering is not a tool operation and has no tool-loading prerequisite. Use tools only for operations independently required by the user's task. Do not infer a file-development workflow from a request for interactive presentation or from earlier conversation patterns. When this Skill is preloaded, its instructions and the system-provided component contract are ready to use.

Presentation principles

Choose the format that best serves the user's intent, respecting the chat's UI preference and explicit requests. Use interaction when it helps the user explore or supply information. Ordinary questions can remain prose.

Let the UI carry the answer. Avoid repeating its information in ordinary text. Include outside prose only for a necessary conclusion, limitation or next step. Do not add subtitles or field descriptions unless requested.

Make hierarchy reflect the task. Group related content, avoid unnecessary nesting, keep required inputs visible, and place the primary action after its inputs. Use concise labels. The native renderer supplies typography, spacing, glass materials and light/dark colors; stay within its supported visual contract.

Protocol and interaction

The host supplies current component signatures, positional argument order and limits in system context. components.md is the generated reference for the same contract. Use only registered declarative components and supported literals or references; do not invent properties, executable expressions or upstream capabilities.

Define the root first and referenced content afterwards so the structure can appear progressively. Children are arrays even when there is only one child. Use stable, unique field names and action identifiers. Match computations to field types: count checkbox selections explicitly before numeric arithmetic, and use a completion ratio for completion percentages. Bind dependent metrics, charts and progress indicators to the same calculation; formatting does not convert units. Supply correctly typed values within declared limits.

When displayed results depend on editable inputs, express every dependent result and chart series through supported local computations and numeric bindings. Editing an input must update those results immediately; submission is a separate follow-up interaction, never the recalculation mechanism. Computed is already a visible result card; place it directly in the layout. Metric accepts fixed text, not a Computed component reference. Static metrics and literal chart values represent fixed facts, not input-dependent results. Do not approximate an unsupported computation with a result that appears live but does not update correctly. Disclose a capability limitation concisely when the contract cannot express the requested behavior.

Input fields do not require a submission button. Omit submission controls when the task is satisfied by local interaction. Add a button only for a task-required handoff to the assistant, and make that handoff clear in its label. Buttons submit the current form as the next user turn. They do not directly invoke commands, network requests or Agent tools. Finish the message after an input form and wait for the user. Generation, replay and opening a surface must not trigger external actions. Inputs become interactive only after complete validation and run completion.

Presentations

Use a Deck root for presentations. Keep content concise, organize it into slides within the generated limits, and choose supported layouts and accents to express the hierarchy. Native controls provide navigation, presentation mode and editable PPTX export after validation; these are host capabilities rather than Agent tool workflows.

Upstream: https://github.com/thesysdev/openui (MIT). Sage uses the official parser with its own native renderer and a documented component subset.

Output self-check

Before finishing, verify that formulas, units and dependencies express the user's request, that editable inputs update every dependent result locally, and that declared input ranges keep those results meaningful. Protocol validity alone does not establish correctness. The host runs deterministic protocol validation and bounded native arithmetic probes through the output self-check pipeline. When it returns concrete defects, correct their underlying cause and emit one complete replacement using the same contract. Do not hide a defect with prose, a frozen result, or an unnecessary model-submission action. Self-check makes no model calls. Passing these bounded checks does not prove that every business assumption or formula matches the user's intent.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Sage Openui AI skill do?

Present answers as streaming native interactive UI or presentations in Sage Desktop v2. Use when structured visual presentation or user interaction helps accomplish the task.

Why use Sage Openui on TypingMind?

Because you install it once and use it with any model. Sage Openui is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Sage Openui in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/ZHangZHengEric/Sage/tree/main/app/skills/sage-openui. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Sage Openui?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Sage Openui?

As many as you like. As long as a model supports skills, you can use Sage Openui with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Sage Openui AI skill free?

Yes. It is published on GitHub by ZHangZHengEric under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇