algorithmic-art

Define an algorithmic philosophy first, then implement it as a p5.js sketch with setup() and draw() functions. Use randomSeed() and noiseSeed() for reproducibility, organize tunable values in a params object, and expose them through slider controls in the HTML viewer.

Seeded randomness initializes p5.js random() and noise() with a fixed seed value so the same seed always produces identical output. This enables Art Blocks-style reproducibility where collectors can share and revisit exact variations by seed number.

Yes, the output is a single self-contained HTML file with p5.js loaded from a CDN and all code inline. It runs directly in any browser or as a claude.ai artifact with no server, build step, or external files required.

Yes, the viewer includes previous, next, random, and jump-to-seed controls for browsing variations. You can generate 100 variations using seeds 1-100, and each seed deterministically reproduces its unique composition.

Copying existing artists' work risks copyright violations. The skill creates original algorithmic philosophies and implementations instead, deriving aesthetic direction from computational concepts like flow fields, recursion, and emergent behavior rather than imitating specific artists.