algorithmic-art · Algorithmic Art

A skill for creating interactive generative art with p5.js and seeded randomness: first an algorithmic aesthetic manifesto (4–6 paragraphs), then expressed as executable code — flow fields, particle systems, noise fields. The beauty lives in the algorithm's execution, not in any single frame.
Output is a single self-contained HTML artifact: seed navigation (previous/next/random/jump), real-time parameter sliders, one-click regenerate and PNG download. Same algorithm, different seeds — different artworks.
Example invocation: "Make an interactive generative piece in an 'organic turbulence' style."
Full brief
Positioning
algorithmic-art is a generative-art creation process: the user's aesthetic intent is first written as an algorithmic philosophy (a computational aesthetic manifesto), then expressed in p5.js as executable code. It produces "living algorithms" — every run, every seed, a unique emergence — never a static image.
Core capabilities
- Algorithmic philosophy writing: a 4–6 paragraph manifesto articulating how computational process, noise and randomness patterns, particle behavior, temporal evolution, and parametric variation express the aesthetic.
- Conceptual seed deduction: distilling a subtle conceptual thread from the original request and weaving it into parameters and emergent behavior rather than literal illustration.
- p5.js implementation: flow fields, particle systems, noise fields, recursive branching, Voronoi relaxation — algorithms built to order from the philosophy, never picked from a pattern menu.
- Seeded randomness: randomSeed / noiseSeed guarantee the same seed always produces the same piece — fully reproducible.
- Interactive artifact: single self-contained HTML file (p5.js via CDN) with seed navigation (previous/next/random/jump-to), parameter sliders, optional color pickers, regenerate, reset, and PNG download.
- Originality first: creates original algorithms rather than copying existing artists' work.
Workflow
- Interpret the user's intent and distill the aesthetic direction
- Write the algorithmic philosophy (4–6 paragraph manifesto)
- Deduce the conceptual seed (a subtle thread woven into parameters and behavior)
- Implement the algorithm in p5.js (starting from the viewer.html template; only the algorithm and parameter sections change)
- Design parameters and UI controls; tune to taste
- Deliver the single-file HTML artifact; the user explores variations via seed navigation
Inputs & outputs
| Input | Notes |
|---|---|
| Aesthetic intent | Required — a sentence, a mood, a reference direction |
| Tunable-parameter preferences | Optional — which dimensions should be adjustable (count/scale/palette…) |
Output: the algorithmic philosophy document (.md) plus the single interactive HTML artifact (.html, opens in any browser).
Boundaries with adjacent skills
- ai-image-gen: text-to-image produces one static image per run; this skill produces an executable generative algorithm — a new piece every run.
Fit
- Creating interactive generative artworks
- Visual experiments needing "one logic, infinite variations"
- Real-time evolving visuals for websites or exhibitions
- Work with clean provenance (original algorithms, nobody's style copied)
Before you start
- No API key needed: p5.js loads from CDN; just open the HTML in a browser.
- The more specific the aesthetic direction (mood, texture, references), the sharper the philosophy; leave it open and the skill improvises freely.
algorithmic-art is part of the Aiglade Skill library. Invoke it from the Aiglade chat box in plain language.