Riso Still logo

Riso Still

Community
sevenevesai
riso-still

Create or improve procedural risograph still art, illustrations, posters, print series and images in this repo. Develops reference-led composition, perspective, pose and ink plates; delivers a native-resolution PNG and self-contained Canvas 2D source. Use for standalone still requests; use riso-film for moving work.

Overview

Publishersevenevesai
Repositoryriso-windowseat
Skill nameriso-still
Stars
303
Forks
32
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 sevenevesai on GitHub. Read the source before you install it.

Installation

Install the Riso Still 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/sevenevesai/riso-windowseat/tree/main/.claude/skills/riso-still
  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/sevenevesai/riso-windowseat.git /tmp/riso-windowseat
mkdir -p .claude/skills
cp -r /tmp/riso-windowseat/.claude/skills/riso-still .claude/skills/riso-still
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Riso Still 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 Riso Still 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 Riso Still 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.

Riso still

Deliver prints/<name>/index.html and a native PNG in out/. Same medium as the films: Canvas 2D only, no libraries, fonts, images or network calls. External references are observation sources, never embedded assets. If the user explicitly wants generated raster imagery, that is a different workflow; do not quietly substitute it. Commands run in tools/.

Develop the picture

Read .claude/rules/riso-plates.md, docs/visual-development.md and docs/drawing.md (its composition and frame-budget sections apply in full); add docs/scene-space.md when perspective or attached details matter and docs/characters.md for figures. Skip film timelines, transitions and sound. examples.md maps the shipped series (prints/cabinet, prints/sceneries, prints/workings); open the matching source before inventing a routine it already has.

Extract subject, feeling, format and exclusions. If the user wants ideas, offer different scenes or viewpoints; if they want art, choose and draw. Default to 1080 square only when no format was given; record other sizes before composing.

  • An open request ("a collection", "detailed sceneries", "your strengths") is permission to art-direct a series of three to five prints under one principle stated up front: one construction style held across different subjects (prints/cabinet), or one technical strength per print (prints/sceneries: constructed perspective with fog, lights with reflection, strata with layered haze).
  • For an unfamiliar subject, look at real references and note proportions, gesture, construction and material; measure in the photo's own pixel frame where it matters (wing outline, vein and eyespot positions). A mood adjective or an unopened image link is not enough.
  • For ambitious work, compare small compositions with different viewpoints and value groupings, then develop the strongest. A local edit needs less setup.
  • Solve support, perspective, occlusion and identity before detail. Draw figure landmarks and contact points before clothing or texture: the hand reaches the tool, the tool meets the work. Build details on shared surfaces so ellipses, slats and legs agree on orientation and scale. Grow natural subjects from their own geometry (a logarithmic spiral, Vogel's floret model, a distance transform from a cavity wall), so every mark sits where growth put it.
  • A film's 0.25 s reading time does not govern a still: sustained detail is allowed, but marks should explain form, material or the moment. Concentrate detail at the focus and leave quiet areas. Random hatching and misregistration do not rescue weak drawing.

Build and inspect

node new-riso.mjs --kind still --out ../prints/<name>/index.html
node verify.mjs ../prints/<name>/index.html --times 0,1,2
node still.mjs ../prints/<name>/index.html --at 0 --out ../out/<name>.png

The starter is blank paper with the print and craft kits and no animation loop. Draw in its art block. A single picture uses duration: 1 and a seek(t) that draws the same picture; a series uses integer print indices (floor(t), as prints/workings does), verified at every index and exported one PNG per index. No MP4, score or artificial motion.

For another native size, change canvas, paper, plate dimensions, projection and screen pitch before baking, and audit square assumptions like W,W. Larger prints must rerasterize geometry and screens at that size; CSS/DPR enlargement and PNG upscaling do not make a higher-resolution original. Press-ready separations need a separate print-production brief.

  • 2160 square (OUT = 2160, K = OUT / W) served both Cabinet and Sceneries. PITCH = 4.6 * K (9.2 device px) is the same picture enlarged, right for landscape masses; small dense subjects took a finer 3.4 * K (6.8 px). Registration scales with K.
  • Fine engraved line needs coverage 1 to print solid; the kit's threshold tile leaves the interstices open even at full coverage (right for film tone), which breaks thin line into dots.
  • A feature under a few screen cells (florets, scales, birds) is modelled by shape and solid overprint, not dot size. A small fixed dither on mid-tones breaks remaining dot-size steps in a gentle ramp; leave paper and solids untouched. At 2160, centre the screen's dots on pixels (docs/drawing.md, tone that prints) or ramps contour into rings.
  • Add debug views by query string before polishing: construction lines alone compared against the reference diagram (it caught over-curved septa in a shell section), and one ink at a time for plate isolation.

Inspect the whole picture at viewing size, a value/silhouette view, and 1:1 crops of difficult anatomy, attachments and screen. Fix perspective, tangencies, missing support or an unclear subject before polishing grain; docs/quality-bar.md lists the review cases. still.mjs exports the canvas backing store and checks repeatability; it does not judge the art. Confirm dimensions and open the PNG.

Keep a short PRINT.md beside substantial work: request, the series principle, references per print, viewpoint, geometry and light decisions, native size and pitch, inspected output and each print's remaining weaknesses. Keep the previous version when revising.

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 Riso Still AI skill do?

Create or improve procedural risograph still art, illustrations, posters, print series and images in this repo. Develops reference-led composition, perspective, pose and ink plates; delivers a native-resolution PNG and self-contained Canvas 2D source. Use for standalone still requests; use riso-film for moving work.

Why use Riso Still on TypingMind?

Because you install it once and use it with any model. Riso Still 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 Riso Still in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/sevenevesai/riso-windowseat/tree/main/.claude/skills/riso-still. 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 Riso Still?

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 Riso Still?

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

Is the Riso Still AI skill free?

Yes. It is published on GitHub by sevenevesai 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 👇