Dirty Pixels

by haorantang97

Dirty Pixels is an Automation skill for Claude Code, published by haorantang97 in LabArt.

282 stars16 forkson haorantang97/LabArtAdded 2026/09/03Repository updated 2026/08/27

View source on GitHub Download repo ZIP

agent-skillschinese-inkcodexcreative-toolsgenerative-art

Install in seconds

Install Dirty Pixels

Copy Dirty Pixels into your Claude Code skills folder. Run the command in your terminal, or review the source on GitHub before installing.

terminal

npx degit https://github.com/haorantang97/LabArt/tree/main/plugins/dirty-pixels/skills/dirty-pixels ~/.claude/skills/dirty-pixels

Requires Node.js. Downloads this skill only — not the rest of the repository — into your Claude Code skills folder.

Without Node.js

git clone https://github.com/haorantang97/LabArt.git

Clones the whole repository, then copy the skill’s own directory into your skills folder yourself.

In this catalog

Source file: plugins/dirty-pixels/skills/dirty-pixels/SKILL.md in haorantang97/LabArt
Installs to: ~/.claude/skills/dirty-pixels
Collection: One of 2 skills cataloged from this repository
Category: Automation — 1754 skills

What Dirty Pixels does

Dirty Pixels重画输入图片为黑底像素粒子画，并可输出静态PNG或无缝循环GIF/MP4。适合把照片或插画转换成带流动笔触和尘化质感的粒子风格。

Dirty Pixels is cataloged under Automation on DirSkills. Dirty Pixels comes from a repository tagged agent-skills, chinese-ink, codex, creative-tools and generative-art.

Documentation

README

Excerpt

Dirty Pixels v5

把一张图重画成黑底像素粒子画。核心脚本 scripts/dirty_pixels.py，确定性（同参数同 seed 逐字节相同），全部美学决定都是参数。给用户交付时必须报出完整命令——那就是版本号。

v4 的「实心网点 mass 层 + 规划蒙版」画法在实践中被否决：蒙版是画出来的几何形状，边界永远贴不住图像；对连续光影塑形的图，mass 层会把层次压成色块。 v5 的正典是纯粒子：不给 --plan，不开 mass 层，每一笔的落点由原图亮度和梯度采样，边界天然贴合。虚空靠 --floor 从亮度上切，不靠蒙版圈。（v4 的 --plan/--mass 仍可用，仅当输入是平实大色块、真需要静止锚定层时才考虑。）

正典配方（与用户逐轮审定的终版）

This is the opening of the README. Read the full README on GitHub.

Frequently asked about Dirty Pixels

What else does haorantang97 publish alongside Dirty Pixels?
Dirty Pixels is one of 2 skills that DirSkills catalogs from haorantang97/LabArt, the repository it ships in. Its siblings there include Antibes Holiday. Each one is a separate skill with its own page in this directory, installs the same way Dirty Pixels does, and is maintained by haorantang97 in that same repository. The rest of the collection is listed on the haorantang97/LabArt page.
How does Dirty Pixels compare to other Automation skills?
Dirty Pixels ranks #1407 by stars among the 1754 Automation skills in this catalog. The most-starred ones next to it are Autonomous Loops, Autonomous Agent Harness and Automation Audit Ops. DirSkills ranks by the star count of the repository each skill ships in, so that order reflects how popular those repositories are rather than any review of Dirty Pixels against them. Open each page to compare what they document and how they install.