mdskills

Learn

Guides, comparisons, and tutorials for AI agent skills.

A pixelated orange character with a hat.
Comparisons

Cursor vs Claude Code: Which Wins for Agentic Work?

Cursor excels at inline IDE flow. Claude Code wins for autonomous multi-file agents. Real pricing, Reddit verdicts, and concrete use cases.

Two computer screens side by side
Comparisons

Claude Code vs Cursor: Which AI Coding Agent in 2026?

An honest comparison of Claude Code and Cursor: skill support, workflows, pricing, and where each one wins.

Server room with cables
Comparisons

Claude Code vs Codex CLI: How They Compare

Claude Code and OpenAI Codex CLI both support SKILL.md. Here is where they differ on skill loading, context, and agent behavior.

Forking paths
Comparisons

SKILL.md vs .cursorrules vs CLAUDE.md: Which Format to Use

Three agent config formats, three different purposes. When to use SKILL.md, when .cursorrules makes sense, and where CLAUDE.md fits.

Whiteboard with diagrams
Comparisons

MCP Servers vs Skills: When to Use Which

MCP servers give agents access to tools and data. Skills give agents instructions and knowledge. Here is when to reach for each.

Digital visualization
Comparisons

Gemini CLI vs Claude Code for AI Agent Skills

Google Gemini CLI now supports SKILL.md. How it compares to Claude Code on skill loading, context handling, and developer experience.

Network connections
Comparisons

MCP vs REST APIs: When to Use Each for AI Agents

MCP and REST APIs both connect agents to external services. Here is when MCP makes sense and when a plain API call is simpler.

AI interface
Comparisons

Claude Code Plugins vs Skills: What Is the Difference?

Plugins and skills both extend Claude Code, but they work differently. When to use each and how they interact.