# Comparisons

> Cursor rules vs skills vs commands vs agents, and AgenticKit vs building it yourself.

Canonical: https://www.agentickit.co/compare
Last updated: 2026-11-18

- [CLAUDE.md vs Skills vs Subagents vs Hooks](https://www.agentickit.co/compare/claude-code-skills-vs-subagents-vs-hooks) - Claude Code's four primitives compared: CLAUDE.md is always-loaded memory, skills are on-demand SKILL.md playbooks, subagents run in their own context window, hooks are deterministic scripts on events. Decision table included.
- [Cursor Rules vs Skills: Official Docs Explained](https://www.agentickit.co/compare/cursor-rules-vs-skills) - What Cursor's own docs say about rules vs skills vs commands, explained in plain language: rules are always-on or glob-scoped constraints, skills are on-demand SKILL.md playbooks, commands are workflows you type with /.
- [Cursor Rules vs Skills](https://www.agentickit.co/compare/cursor-rules-vs-skills-decision) - Pick between a Cursor rule and a skill with a 4-question decision tree: always-on vs on-demand, short ban vs long playbook, path-scoped vs task-scoped.
- [Cursor Rules vs AGENTS.md vs CLAUDE.md](https://www.agentickit.co/compare/cursor-rules-vs-agents-md-vs-claude-md) - Which config file Cursor, Claude Code, Codex and Windsurf read, how to keep one source of truth, and when to convert between .mdc, AGENTS.md and CLAUDE.md.
- [Cursor Skills vs Cursor Commands](https://www.agentickit.co/compare/cursor-skills-vs-commands) - Skills are playbooks the agent may load. Commands are workflows you type with /. How /seo-audit and the on-page SEO skill work together.
- [Cursor Agents vs Cursor Commands](https://www.agentickit.co/compare/cursor-agents-vs-commands) - Agents are roles. Commands are workflows that adopt those roles. How /ship uses tech-lead without forking 34 processes.
- [Cursor Commands vs Hooks](https://www.agentickit.co/compare/cursor-commands-vs-hooks) - Commands are prompts you type with /. Hooks are scripts that run on agent events. A decision table plus three setups: format on edit, block dangerous commands, audit log.
- [Cursor Commands vs Rules Files](https://www.agentickit.co/compare/cursor-commands-vs-rules) - A command in Cursor is a markdown prompt you trigger with /. A rules file is always-on or glob-scoped context the model reads on every turn. When each one is the right tool.
- [AgenticKit vs DIY Cursor Setup](https://www.agentickit.co/compare/agentickit-vs-diy-cursor) - Writing your own SKILL.md files vs installing 46 agents, 61 skills, and 47 commands. Time, drift, and when DIY is the right call.
- [AgenticKit vs Cursor Background Agents](https://www.agentickit.co/compare/agentickit-vs-cursor-background-agents) - Background Agents run Cursor's own agent in a remote cloud sandbox so it can work without tying up your editor. AgenticKit is a set of role, skill, and command files that live in your repo. They solve different problems and most teams want both.
- [Best Cursor Skills](https://www.agentickit.co/compare/best-cursor-skills) - The Cursor skills we actually ship with: Next.js, Stripe, on-page SEO, positioning, and project conventions. Not a scraped marketplace list.
- [Best Cursor Commands](https://www.agentickit.co/compare/best-cursor-commands) - The Cursor slash commands that actually ship SaaS: /ship, /fix, /audit, /positioning, /launch, /seo-audit. When not to use each.
- [Best Cursor Agents](https://www.agentickit.co/compare/best-cursor-agents) - Named Cursor agent roles, not the built-in Agent product. tech-lead, seo-specialist, launch-coordinator, security-auditor, and when to invoke each.
- [Cursor Skills vs Claude Code Skills](https://www.agentickit.co/compare/cursor-skills-vs-claude-skills) - Cursor skills and Claude Code skills both ship as a SKILL.md with YAML frontmatter, but they live in different folders, load on different triggers, and don't move between tools without edits. Side-by-side comparison and a verdict for solo founders.
- [Claude Code Skills vs Cursor Skills: Storage, Portability, and Which Wins for SaaS Teams](https://www.agentickit.co/compare/claude-code-skills-vs-cursor-skills-storage-portability) - Claude Code skills can live in a personal, cross-project folder (~/.claude/skills/) or a project folder (.claude/skills/). Cursor only has the project folder (.cursor/skills/) — there is no personal skills directory yet. What that gap means for a solo SaaS founder running both tools.
- [Cursor Skills vs MCP Servers](https://www.agentickit.co/compare/cursor-skills-vs-mcp) - A skill is instructions loaded on demand. An MCP server is a live tool and data connection. Decision table, context cost, a Stripe example done both ways, and the hybrid pattern.
- [Cursor Rules for Next.js App Router Projects](https://www.agentickit.co/compare/cursor-rules-nextjs-app-router) - A short, glob-scoped .cursor/rules file for App Router repos: server/client boundary, Server Actions, Suspense, and the parts that belong in a skill instead.
- [Cursor Rules for Stripe & Billing Codebases](https://www.agentickit.co/compare/cursor-rules-for-stripe) - Four glob-scoped Cursor rules for Stripe-heavy repos: webhook verification, idempotent handlers, subscription state, and secret hygiene. Copy-paste .mdc files plus when to graduate a rule into the Stripe billing skill.
- [Claude Code Plugins vs Skills vs MCP](https://www.agentickit.co/compare/claude-code-plugins-vs-skills-vs-mcp) - Plugins bundle and distribute, skills teach a playbook on demand, MCP servers connect to live external systems. A decision table and three worked scenarios for when to use each in Claude Code.
- [Cursor Rules Examples: React Native, Go, Flutter, Svelte](https://www.agentickit.co/compare/cursor-rules-examples-react-native-go-flutter-svelte) - Annotated .cursor/rules/*.mdc starter files for four stacks Cursor's own docs barely cover: React Native, Go, Flutter, and Svelte — plus the mistakes that show up when a rule gets ported from one framework to another.
- [Claude Code Agents vs Claude Code Commands](https://www.agentickit.co/compare/claude-code-agents-vs-commands) - A Claude Code subagent (.claude/agents/*.md) is a role with its own context window and tool grant. A slash command (.claude/commands/*.md) is a prompt you trigger with /. How /ship invokes tech-lead without forking 34 sessions, plus a decision table.
- [Claude Code Role Agents vs Cursor Custom Agents](https://www.agentickit.co/compare/claude-code-role-agents-vs-cursor-custom-agents) - Looking for a Claude Code SEO agent, growth agent, support-triage agent, or QA agent? Each is a .claude/agents/<name>.md role with its own tools and model. What each one does, which skills it's wired to, and how installing it differs from Cursor's @-mentioned agent files.
- [Cursor Rules vs Claude Code CLAUDE.md](https://www.agentickit.co/compare/cursor-rules-vs-claude-md) - How Cursor's .cursor/rules/*.mdc system differs from Claude Code's CLAUDE.md — frontmatter, path scoping, always-apply flags — and the actual steps to migrate rules into a CLAUDE.md.
- [Claude Code Hooks vs MCP Servers](https://www.agentickit.co/compare/claude-code-hooks-vs-mcp-servers) - Hooks gate and automate lifecycle events — lint, format, block a dangerous command. MCP servers give the agent live access to external systems — Stripe, Supabase, GitHub. A decision table for which to reach for first, and when to use both.
- [Claude Code Plugins vs Commands](https://www.agentickit.co/compare/claude-code-plugins-vs-commands) - A plugin bundles commands, skills, agents and hooks into one shareable, versioned install. A lone command is a single markdown file for one project. A checklist for deciding which to build, plus how to migrate a /ship command into a plugin.
- [Cursor Commands vs Claude Code Commands](https://www.agentickit.co/compare/cursor-commands-vs-claude-code-commands) - Cursor's .cursor/commands/*.md and Claude Code's .claude/commands/*.md are both markdown prompts triggered with /, but they store, distribute, and convert differently. A decision table plus the recommendation for solo founders picking one primary tool.
- [Claude Code Skills vs Cursor Rules vs Codex Skills](https://www.agentickit.co/compare/claude-code-skills-vs-cursor-rules-vs-codex-skills) - Three on-demand instruction formats, three different load triggers: Claude Code's SKILL.md, Cursor's .cursorrules/.mdc, and Codex's native skills. What each file looks like, how portable it is, and which one to standardize on if you run more than one tool.
- [Sentry MCP + Claude Code: The Agent That Triages Your Errors](https://www.agentickit.co/compare/sentry-claude-code-agent) - How to wire the Sentry MCP server into Claude Code so the debugger subagent can pull a live stack trace, find the failing commit, and propose a fix — the mcp.json config, which agent calls it, and a full auto-triage walkthrough.
- [Claude Code vs ClaudeMarketplaces](https://www.agentickit.co/compare/claude-code-vs-claudemarketplaces) - ClaudeMarketplaces.com lists individual community Claude Code skills browsable by GitHub author and repo, like affaan-m's team-builder skill. AgenticKit installs a pre-wired 46-agent, 61-skill, 47-command team in one command. Side-by-side on setup time, maintenance, and when a single marketplace skill is actually the better pick.
- [Claude Code Skills vs Claude Agent Skills (API)](https://www.agentickit.co/compare/claude-code-skills-vs-agent-sdk-skills) - Three different things get called 'Claude skills': Claude.ai consumer skills, Claude Code repo skills, and Claude Agent SDK skills. A decision table for which one you're actually building, plus why a plugin can silently override a repo skill.
- [Claude Code Subagent Limits: Concurrency and the Model Setting](https://www.agentickit.co/compare/claude-code-subagent-concurrency-model-setting) - How many Claude Code subagents can run at once, how to set a different model per subagent in frontmatter, and why subagents don't pollute the main thread's context. Reconciles the conflicting '10 at once' claim floating around.
- [Claude Code Hooks vs Plugins](https://www.agentickit.co/compare/claude-code-hooks-vs-plugins) - A hook is a deterministic script enforcing a rule at runtime; a plugin is the versioned, shareable bundle that packages hooks (plus commands, skills, agents) for distribution. A decision table and the migration path from a loose hook script to a packaged plugin.
- [Claude Code Agent Teams vs Cursor Background Agents](https://www.agentickit.co/compare/claude-code-agent-teams-vs-cursor-background-agents) - Claude Code's experimental agent teams run multiple subagents in parallel inside one session on your machine; Cursor Background Agents run one agent per task in a remote cloud VM. Different coordination models, different setup, different token cost. Side-by-side breakdown.
- [Claude Code Skills vs Plugins: Token Cost Breakdown](https://www.agentickit.co/compare/claude-code-skills-vs-plugins-token-cost) - Skills and plugins cost tokens at different layers — discovery, frontmatter, and invocation. A line-by-line cost breakdown for both, with the cases where the difference actually matters on a solo repo's context budget.
- [Cursor MCP Setup vs Claude Code MCP Setup](https://www.agentickit.co/compare/cursor-mcp-setup-vs-claude-code-mcp-setup) - Cursor configures MCP servers by hand-editing mcp.json; Claude Code adds them with claude mcp add from the CLI. Config file locations, project-vs-global scoping, servers that behave identically, and a migration checklist for founders running both tools.
