A collection of Golang agentic skills designed to enhance AI coding assistants with domain-specific expertise. These skills are reusable instruction sets for areas like language, testing, security, and observability, loaded on demand to avoid bloating the AI's context. The project targets developers using AI coding assistants, aiming to provide production-ready Go project guidance and improve code quality.

How It Works

This project offers atomic, cross-referencing skills that integrate with various AI coding tools (e.g., Claude Code, Gemini CLI, Copilot). Skills are triggered via descriptions in SKILL.md frontmatter and leverage lazy loading for detailed content from referenced markdown files, minimizing initial context overhead. This approach ensures that only relevant expertise is loaded when needed.

Quick Start & Requirements

• Primary Install: Use the skills CLI: npx skills add https://github.com/samber/cc-skills-golang --skill '*'. Integration methods for Claude Code, Openclaw, Gemini CLI, Cursor, Copilot, OpenCode, Codex, and Antigravity are also provided via git clone or plugin commands.
• Prerequisites: Requires an AI agent tool compatible with the skills format. No other non-default prerequisites are specified.
• Resource Footprint: Recommended skills load approximately 1,100 tokens at startup, with full content loaded only on demand.
• Links: Repository: https://github.com/samber/cc-skills-golang

Highlighted Details

• Skills are explicitly stated as human-curated ("Edited, tested, reviewed and reworked by a human. No AI slop here.").
• Claims significant "Error rate gap" reductions across various categories, e.g., golang-safety at -58% and golang-modernize at -61%.
• Skills are designed to be atomic and cross-referenced, with minimal overlap.
• Features flags like ⚡ (Command available), 🧠 (Ultrathink automatically), and ⚙️ (Overridable) indicate specific functionalities.

Maintenance & Community

The project is primarily authored by Samuel Berthe. The README does not provide links to community channels (like Discord/Slack), social handles, or a public roadmap.

Licensing & Compatibility

• License: MIT License.
• Compatibility: The MIT license is permissive and generally suitable for commercial use and integration into closed-source projects without significant restrictions.

Limitations & Caveats

The author emphasizes human curation over AI-generated content. There is an estimated 50% overlap between golang-naming/golang-code-style skills and linters like golangci-lint. Some skills are marked as "Work in progress" (👷) or "To-do" (❌). Customization may be required if a team has specific conventions, using the override mechanism.