Glyph Kit logo

Glyph Kit

Community
alextraykov
glyph-kit

Draws product-design figures and one-page readouts, dashboards and status updates as text with block (░▒▓█) and box-drawing characters - research and survey results, funnels and A/B tests, UX flows and state machines, roadmaps, screen wireframes, design-system docs, handoff and QA checklists, motion specs and moving diagrams. Use for an ASCII or Unicode diagram, text wireframe, sparkline or bar chart in Markdown, a PR, a ticket or a doc; a design review deck; a report or audit as a PDF; or turning a text wireframe into an HTML prototype or components from the project's design system. Draws from a small JSON spec with real, sourced data and exports Markdown, MDX, HTML, SVG, PNG and PDF.

Overview

Publisheralextraykov
Repositoryglyph-kit
Skill nameglyph-kit
Stars
77
Forks
0
Bundled files
87
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.

  • 87 bundled files

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

  • Open source

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

Installation

Install the Glyph Kit 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/alextraykov/glyph-kit/tree/main/skills/glyph-kit
  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/alextraykov/glyph-kit.git /tmp/glyph-kit
mkdir -p .claude/skills
cp -r /tmp/glyph-kit/skills/glyph-kit .claude/skills/glyph-kit
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Glyph Kit 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 Glyph Kit 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 Glyph Kit 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.

Glyph kit

You write a spec, a small JSON object; a script with no dependencies draws it. The kit lays out every figure, so its output always lines up: never draw a figure by hand when a type covers it.

Work in the user's folder, not this skill's, so specs and outputs land there. <skill> below is this skill's folder; file paths in this guide are relative to it.

Make one

  1. Pick the type from this table; it goes in the spec's kind field.

    The task is aboutTypes
    Interviews, surveys, usability tests, quoteslikert themeMatrix affinity quoteCard taskStats emotionCurve
    Funnels, retention, A/B results, metrics, chartsfunnel cohort abResult metricPanel barChart compare forecast meter sparkline heatmap calendarStrip
    Flows with decisions, state machines, journeys, sitemapsflow branchFlow stateMachine journeyMap tree
    Roadmaps, priorities, ownership, timelines, decisionsboard impactEffort swimlane gantt statusKey decisionLog
    Screens: one screen, an app shell, a card grid, list and detail, a modalscreen responsive appShell cardGrid listDetail modal
    Tokens, scales, colour, contrast, component states, layersspacingScale typeScale palette colourRoles contrast stateMatrix layerStack tokenTable table
    Redlines, QA checklists, component anatomy, before and afterredline qaChecklist anatomy compare
    Motion intent, timing, transitions, stagger, easingmotionTiming intentCard transitionSpec stagger filmstrip easingTable
    Work moving over time: rework loops, state machines, queues, rollouts, loadingloop states queue rollout loading
    A whole deliverable on one page: readout, dashboard, PR summary, design review, status updatereadout dashboard prSummary designReview statusUpdate

    Unsure between two? Read the area's guide, reference/areas/<area>.md (research, metrics, flows, planning, wireframes, systems, handoff, motion, process, layouts): it shows each type drawn.

  2. Get its data shape: node <skill>/scripts/glyph-kit.mjs guide <kind> prints the kind's example spec and what to watch for. Copy that shape; don't read the scripts.

  3. Write the spec as JSON next to the document that uses it:

    json
    {
      "kind": "funnel",
      "title": "Signup to paid, last 30 days",
      "intent": "Find the step where most people leave.",
      "source": "Product analytics export, 2026-10-15",
      "data": [{ "label": "Visited pricing", "value": 12400 }, { "label": "Paid", "value": 610 }]
    }

    Data comes from the user or their files ("data": "@results.csv" reads a file next to the spec), never from you. If numbers are missing, ask. Only if the user wants a sketch anyway, start source with "Illustrative": every format then carries a SAMPLE DATA stamp. A spec with numbers and no source is refused.

  4. Render, fix, render again in one command:

    bash
    node <skill>/scripts/glyph-kit.mjs render spec.json --to markdown > figure.md

    The kit refuses a spec it can't draw truthfully and says why in a sentence. Labels over the limit (usually 24 cells) are cut with "…", and the kit lists each one on stderr with its limit. Shorten those labels and render again: a figure with "…" loses words.

  5. Put the conclusion in the sentence around the figure. The figure supports it.

Kit output needs no further checking: frames close and widths fit by construction. When a spec is refused, run check spec.json to see every problem at once instead of fixing them one draw at a time. Run node <skill>/scripts/lint.mjs <file.md> only on figures drawn by hand.

Other outputs

Every command is node <skill>/scripts/glyph-kit.mjs <command>.

OutputCommandRead first
Text, Markdownrender spec.json [--to text|markdown]nothing
HTML page; a screen becomes a working prototyperender spec.json --to html [--map design-system.map.json]reference/fidelity.md for prototypes and code
Deckdeck deck.json > deck.htmlguide deck
Report (cover, summary, sections)export report.json --to html|pdfguide report
MDX, SVG, PNG, PDF of anythingexport <file> --to md|mdx|html|svg|png|pdfreference/export.md
Every problem with a spec at oncecheck spec.json or check notes.mdnothing
A figure that stays editable in Markdownrender spec.json --to markdown --with-spec, then redraw notes.md after a spec change or read notes.md after a drawing editreference/editing.md
Your colours, or a palette--theme ink, --theme theme.json or --theme mixed:#7a5cff; roles from your tokens with --mapreference/colour.md
Moving diagram: a process over timelive spec.json --to html (or --to markdown for its still)reference/live.md
Several layouts as one documentexport layouts.json --to html|pdf, a layout to a pagereference/areas/layouts.md

Decks and reports open with the kit's logo: ░▒▓█ side by side, in that order and in one colour. Keep it that way when you edit their HTML or draw a title card or still yourself.

HTML outputs are self-contained (font, styles and, for moving pages, the player inlined): 40–200 KB. Don't read them back to check them; render the same spec --to text instead. PDF and PNG use a Chromium browser already on the machine; everything else needs nothing.

When another agent or a later turn needs the figure, pass the spec (a few hundred bytes), not the drawing. If a drawing must go into a model's context, add --compact (plain ASCII, same layout).

Rules

  1. Every glyph has a referent: a number, a state or a structure. No decoration, no green-on-black or glyph rain unless the product is a terminal.
  2. One meaning per glyph per document. █ measured or done, ░ projected or the empty part of a scale, · an empty cell.
  3. To call something out, use focus, never a colour. "Highlight the drop-off", "show where it fails": name it as drawn in focus.on, pick a tone, and say the point in focus.say (reference/spec.md). It colours, recedes the rest and arrives last in every form, and the text keeps a ► line.
  4. Numbers are printed as text next to the glyphs, and every figure with numbers shows its source.
  5. Labels may use any script. Width is counted in cells: CJK takes two. Set lang for casing and number format. Font and right-to-left limits: reference/characters.md.
  6. Every type moves in HTML and decks. The processes and a test result with readings over time keep moving; every other type arrives once, in the order that carries its point, then stays still. Each keeps a still that makes the same point; it is what reduced motion, print and Markdown show. Say what the motion shows in the sentence next to it.
  7. Drawing by hand (only when no type fits): use the glyphs in reference/characters.md, follow reference/layout.md (two cells of padding inside a box, four between boxes, at most 80 columns), then lint.
  8. Going up in fidelity (prototype, production code): use only the components and tokens the design-system map names, and ask when it has no entry. reference/spec.md describes the spec fields and the map.

If figures look misaligned where the user reads them, the font is drawing a glyph at the wrong width: suggest Departure Mono (free, departuremono.com), never require it.

Bundled files

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

and 27 more files.

Frequently asked questions

What does the Glyph Kit AI skill do?

Draws product-design figures and one-page readouts, dashboards and status updates as text with block (░▒▓█) and box-drawing characters - research and survey results, funnels and A/B tests, UX flows and state machines, roadmaps, screen wireframes, design-system docs, handoff and QA checklists, motion specs and moving diagrams. Use for an ASCII or Unicode diagram, text wireframe, sparkline or bar chart in Markdown, a PR, a ticket or a doc; a design review deck; a report or audit as a PDF; or turning a text wireframe into an HTML prototype or components from the project's design system. Draws...

Why use Glyph Kit on TypingMind?

Because you install it once and use it with any model. Glyph Kit 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 Glyph Kit in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/alextraykov/glyph-kit/tree/main/skills/glyph-kit. 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 Glyph Kit?

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 Glyph Kit?

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

Is the Glyph Kit AI skill free?

Yes. It is published on GitHub by alextraykov 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 👇