kaankiziltug/logo-design-skill

What it is

logo-design-skill is a ready-made "skill" that turns AI coding agents (Claude, Gemini CLI, Codex CLI, Cursor, etc.) into professional logo designers. It packages a step-by-step design process, a searchable library of 1,400+ real-world SVG logos, and a set of Python tools that audit and test logo drafts — all in the open Agent Skills format (a folder with a SKILL.md).

It's not a standalone app; it's an add-on that gives an AI agent the knowledge and tooling to go from a client brief to production-ready SVG files and brand guidelines.

How it works

The skill follows a structured workflow:

1. Brief & research – The agent asks clarifying questions and studies category conventions using the logo library.
2. Concepting – It generates 8–12 one-line ideas, then builds 3 in SVG.
3. Testing & refinement – Python scripts audit each SVG for craft issues (e.g., off-angle lines, tiny details, complexity) and generate test sheets that check readability at 16px, one-colour versions, contrast, and how the logo looks on a shelf next to competitors.
4. Checkpoint – The agent shows the concepts as one overview image with a recommendation and stops. Nothing else is produced until the user picks a direction.
5. Delivery – Once approved, the agent produces the full kit: colour palette, lockups, presentation boards with industry mockups, favicon/app-icon set, and brand guidelines.

Key components

• Design principles & process – Embedded in SKILL.md: discovery, word mapping, mark types, geometric construction, optical corrections (overshoot, bone effect), colour, typography, testing, etc.
• Reference library – 1,400+ SVG logos, each classified by mark type, technique, geometry, subject, typography, mood, and industry. Searchable from the command line (search_library.py) and browsable in a local gallery. Used to study construction and avoid look-alikes — never to copy.
• Python tools (dependency-free):
  • svg_audit.py – checks an SVG against logo principles (complexity, angles, tiny details, contrast) and gives a production-readiness score.
  • concept_sheet.py – builds overview sheets of multiple concepts.
  • preview_sheet.py – creates test sheets (16px pixel test, one-colour, reversed, mirror, competitor shelf test).
  • render_png.py – renders SVGs to PNG.
  • package_skill.py – packages the skill for upload to Claude.ai.

Use cases

• Ask your AI agent to design a logo, wordmark, monogram, app icon, or favicon.
• Critique an existing logo and suggest a refresh.
• Generate multiple directions and test them.
• Produce a full brand kit (colour, lockups, mockups, guidelines).

Tech notes

• Uses the open Agent Skills format, so it works with any agent that supports skills (Claude Code, Gemini CLI, Codex CLI, Cursor, GitHub Copilot).
• The tools are plain Python 3, no external dependencies.
• The skill renders its own drafts to PNG and checks them visually, so it works best with models that can view images.

Bottom line

This is a practical, well-crafted project that extends AI agents with a specialised creative capability. It's not a toy — it includes a real design methodology, a large reference dataset, and automated quality checks, all packaged for easy integration into modern AI coding tools.