agentsclimarketplace

Generative art deployment

Skill organvm-iv-taxis/a-i--skills/skills/creative/generative-art-deployment

Deploy generative art projects for exhibition, web galleries, and print production. Covers rendering pipelines, resolution management, gallery hosting, and archival strategies for algorithmic artworks. Triggers on generative art deployment, art exhibition setup, or digital art publishing requests.From its SKILL.md

Install
npx -y skills add organvm-iv-taxis/a-i--skills --skill generative-art-deployment

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 17 stars17 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its file declares

Copied from the file, not written here

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

7.1 KB, as published. Nobody here has run it

Generative Art Deployment

Move generative artworks from development to exhibition, print, and archive.

Deployment Targets

TargetFormatResolutionConsiderations
Web galleryHTML/JS, WebGLScreen (72-96 DPI)Performance, loading time
Physical printPNG/TIFF300 DPI minimumColor profiles, bleed
LED installationVideo/WebGLPanel-specificBrightness, refresh rate
NFT/on-chainPNG/SVG/HTMLVariableFile size, determinism
Social mediaPNG/MP4Platform-specificCompression, aspect ratio
ArchiveSource + rendersMaximumReproducibility

Web Gallery Deployment

Static Gallery Structure

gallery/
├── index.html          # Gallery grid/navigation
├── works/
│   ├── piece-001/
│   │   ├── index.html  # Full-screen viewer
│   │   ├── sketch.js   # Live generative code
│   │   ├── thumbnail.png
│   │   └── metadata.json
│   └── piece-002/
│       └── ...
├── assets/
│   ├── style.css
│   └── gallery.js
└── catalog.json        # Machine-readable catalog

Work Metadata

{
  "title": "Recursive Bloom #47",
  "artist": "Artist Name",
  "date": "2026-03-20",
  "medium": "Generative, p5.js",
  "dimensions": "3840 × 2160",
  "seed": 1742518400,
  "parameters": {
    "complexity": 0.7,
    "palette": "autumn",
    "iterations": 5000
  },
  "description": "An exploration of recursive growth patterns...",
  "series": "Recursive Bloom",
  "edition": "1/1",
  "tags": ["recursion", "organic", "growth"]
}

Performance Optimization

// Render once, display static
function setup() {
  const canvas = createCanvas(3840, 2160);
  noLoop(); // Don't animate
}

function draw() {
  randomSeed(SEED);
  // ... generate artwork
  saveCanvas('output', 'png');
}

// For interactive pieces: use requestAnimationFrame
// with quality degradation on low-end devices
function draw() {
  if (frameRate() < 30) {
    reduceComplexity();
  }
}

Print Production

Resolution Pipeline

def render_for_print(sketch_path: str, width_inches: float, height_inches: float, dpi: int = 300):
    pixel_width = int(width_inches * dpi)
    pixel_height = int(height_inches * dpi)

    # Add bleed (0.125 inches on each side)
    bleed = int(0.125 * dpi)
    total_width = pixel_width + 2 * bleed
    total_height = pixel_height + 2 * bleed

    return {
        "canvas_width": total_width,
        "canvas_height": total_height,
        "safe_area": {
            "x": bleed, "y": bleed,
            "width": pixel_width, "height": pixel_height,
        },
        "dpi": dpi,
        "format": "TIFF",  # Lossless for print
        "color_profile": "sRGB",  # Or Adobe RGB for wide gamut
    }

Color Management

from PIL import Image, ImageCms

def convert_for_print(input_path: str, output_path: str):
    img = Image.open(input_path)
    srgb_profile = ImageCms.createProfile("sRGB")
    # For fine art printing, embed the ICC profile
    img.save(output_path, "TIFF", dpi=(300, 300), icc_profile=ImageCms.ImageCmsProfile(srgb_profile).tobytes())

Print Sizes

SizeInchesPixels (300 DPI)
A48.3 × 11.72490 × 3510
A311.7 × 16.53510 × 4950
A216.5 × 23.44950 × 7020
24×36 poster24 × 367200 × 10800

Rendering Pipeline

Batch Rendering

import subprocess
from pathlib import Path

def batch_render(sketch: str, seeds: list[int], output_dir: str, width: int, height: int):
    Path(output_dir).mkdir(parents=True, exist_ok=True)
    for seed in seeds:
        output = f"{output_dir}/render_{seed:08d}.png"
        subprocess.run([
            "node", sketch,
            "--seed", str(seed),
            "--width", str(width),
            "--height", str(height),
            "--output", output,
        ], check=True)

# Render a series
batch_render("sketch.js", seeds=range(1, 101), output_dir="renders/series-01", width=3840, height=2160)

Deterministic Rendering

// Ensure reproducibility: same seed = same output
function setup() {
  const seed = parseInt(getURLParam('seed') || '42');
  randomSeed(seed);
  noiseSeed(seed);
  // Record seed in metadata
  document.title = `Piece #${seed}`;
}

Exhibition Patterns

Installation Checklist

  • Hardware specs confirmed (GPU, resolution, orientation)
  • Autostart on boot configured
  • Crash recovery (watchdog/supervisor process)
  • No UI chrome (cursor hidden, fullscreen)
  • Network not required (all assets local)
  • Power failure recovery tested

Kiosk Mode Setup

# Linux kiosk mode
#!/usr/bin/env bash
xset -dpms          # Disable power management
xset s off           # Disable screen saver
unclutter -idle 0.5 & # Hide cursor

chromium-browser \
  --kiosk \
  --disable-infobars \
  --disable-session-crashed-bubble \
  --noerrdialogs \
  file:///home/gallery/piece/index.html

Archival Strategy

Archive Package

archive/
├── README.md           # How to run this piece
├── source/             # Original source code
│   ├── sketch.js
│   └── package.json
├── renders/            # High-res rendered outputs
│   ├── render_001.tiff
│   └── render_001.png
├── metadata.json       # Full metadata including parameters
├── dependencies/       # Vendored dependencies
│   └── p5.min.js
└── documentation/
    ├── process.md      # Artist statement, process notes
    └── screenshots/    # Exhibition documentation

Reproducibility Contract

{
  "runtime": "node 20.x + p5.js 1.9.x",
  "seed": 42,
  "canvas": "3840x2160",
  "parameters": {},
  "checksum": "sha256:abc123...",
  "rendered": "2026-03-20T10:00:00Z"
}

Anti-Patterns

  • No seed recorded — Every generative piece must have a reproducible seed
  • Web-only renders — Always produce high-res static renders for print/archive
  • Missing metadata — Parameters, dimensions, and creation date are essential
  • No offline fallback — Exhibition pieces must work without network
  • Lossy-only archives — Keep lossless TIFF/PNG alongside compressed versions
  • Undocumented dependencies — Vendor or lockfile all runtime dependencies

Keep looking

Skills are one crate of 608,253. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.