kaankiziltug/logo-design-skill

A comprehensive logo-design skill for Claude, Gemini CLI, Codex and other AI agents: principles, process, SVG craft, testing tools and a 1,400+ logo reference library.

What it is

logo‑design‑skill is a plug‑in (called a skill) that lets Claude, Gemini, Codex, Cursor, GitHub Copilot or any other LLM‑agent that understands the Agent Skills format act as a full‑featured logo designer. The skill contains a description, a set of Python utilities and a curated library of ~1 400 real‑world SVG logos. When an agent runs the skill it follows a disciplined design workflow – from brief to concept sketches, audits, tests and finally a production‑ready brand‑kit (SVG logo, favicons, mock‑up boards, style guide, etc.).

How it works

  1. Brief – The user supplies a short brief (industry, tone, constraints). The skill asks follow‑up questions if needed.
  2. Research – It searches the built‑in SVG library for similar marks, extracting visual patterns (mark type, geometry, colour, typography, mood, industry).
  3. Concept generation – The agent drafts 8‑12 one‑line ideas, then creates 3 SVG concepts.
  4. Audit & test – Each concept is run through a set of Python checks:
    • geometric correctness (overshoot, bone effect, exact angles)
    • pixel‑size legibility (16 px, 32 px, 64 px)
    • one‑colour, reversed, squint, mirror, context and competitor‑shelf tests
  5. Checkpoint – The skill shows a greyscale overview image with a recommendation and lets the user either accept a direction or ask for revisions.
  6. Kit creation – Once a direction is chosen, the skill produces the full brand kit: coloured SVG, PNG renders, favicons, web‑manifest icons, lock‑up variations, and a presentation board with industry‑specific mock‑ups.

Who might use it

  • Design‑focused AI agents – Claude Code, Gemini CLI, Codex CLI, Cursor, Copilot, etc., that can load an Agent Skill folder.
  • Design teams – to speed up the early‑stage logo ideation and ensure every concept passes basic visual‑legibility tests.
  • Developers building AI‑driven design tools – the skill’s Python utilities can be called directly for custom pipelines.
  • Educators / students – the library of 1 400+ classified SVG logos is a handy reference for learning logo construction.

What’s inside the repo

Item Description
skills/logo-design/ The skill definition (SKILL.md, prompts, metadata) that agents read.
tools/ Dependency‑free Python scripts for SVG audit, test‑sheet generation, PNG rendering, favicon set creation, and presentation‑board assembly.
library/ 1 400+ real‑world SVG logos, each tagged by mark type, geometry, industry, mood, etc.; searchable via the CLI.
docs/ Images used in the README (hero, example runs, mock‑up boards).
tests/ Automated tests that verify the Python tools work and that the skill’s workflow produces the expected files.
package_skill.py Helper script that bundles the skill (with or without the large SVG library) into a zip for Claude Desktop upload.

Installation & getting started

  1. Claude Code (recommended) – add the skill from the plugin marketplace with:
    /plugin marketplace add kaankiziltug/logo-design-skill
/plugin install logo-design@logo-design-skill
  2. Manual install – clone the repo and copy skills/logo-design into the appropriate ~/.claude/skills/ (or the equivalent folder for Gemini, Codex, etc.).
  3. Claude Desktop / other agents – download the zip from the Releases page (or run python3 tools/package_skill.py) and upload it via the Settings → Capabilities → Skills UI.
  4. Run – start a new chat with the agent and ask for a logo. The agent will automatically load the skill, ask the brief questions, and walk you through the checkpoint → kit flow.

Example output

The README shows 28 end‑to‑end runs (e.g., Kiln coffee roaster, Zestly food‑delivery app, Keyfort password manager). For each brand the skill:

  • presents three greyscale concepts at 16 px/32 px/64 px,
  • highlights any geometric or contrast issues it found,
  • recommends one concept, and
  • after acceptance delivers a full SVG logo, PNG previews, a favicon set, and a presentation board with industry‑specific mock‑ups.

Why it matters

  • Consistent, repeatable design – The skill encodes classic logo‑design principles (optical overshoot, contrast ratios, grid alignment) into automated checks.
  • Leverages LLM creativity – The agent can explore many concept ideas quickly, while the skill keeps the output grounded in proven visual rules.
  • Reusable asset library – Designers can browse a large, classified collection of real logos directly from the CLI, helping avoid accidental look‑alikes.
  • Agent‑agnostic – Because it follows the open Agent Skills spec, the same skill works across Claude, Gemini, Codex, Cursor, etc., making it a portable building block for any AI‑driven design workflow.

TL;DR

logo-design-skill turns any LLM agent that supports the Agent‑Skills format into a disciplined logo‑designer. It guides the agent through brief → concept → audit → checkpoint → full brand‑kit, using a Python‑only toolset and a 1 400‑plus SVG reference library. Install via Claude’s plugin marketplace or by copying the skill folder, then ask the agent for a logo and let the skill handle the heavy lifting.

Related

  • Project
    Owl-Listener/designer-skills

    A set of Claude Code/Gemini‑CLI plugins that turn design knowledge into machine‑readable skills and ready‑to‑run commands, letting AI agents execute full design workflows (research, strategy, UI, interaction, systems, ops, critique) directly from the assistant.

  • Project
    Yuan1z0825/nature-skills

    A library of 22 packaged AI-agent workflows ("skills") for academic research — paper reading, Nature-style writing, figure generation, reviewer responses, literature search, and patent drafting — installable into Claude Code, Codex, Chatbox, and other agents.

  • Project
    OneWave-AI/claude-skills

    A library of over 230 standardized, portable skill files that extend the capabilities of AI coding agents like Claude Code and Cursor for complex engineering, design, and business tasks.

  • Project
    msitarzewski/agency-agents

    A library of markdown‑based AI agent personas (e.g., Frontend Developer, AI Engineer, UI Designer) that can be installed into code‑assistant tools like Claude Code, Cursor, Gemini CLI, etc., via a native desktop app or command‑line scripts. Each agent defines a role, workflow, deliverables and success metrics, enabling consistent, specialist‑level behavior from LLMs.

  • Project
    dzhng/skills

    A library of reusable, composable AI‑agent “skills” (prompt‑driven actions) for building autonomous software‑factory loops, installable via npm and usable with many coding assistants.