Anthropic · skills, agents, hooks, and plugins

Best Claude Coderules & configs — RuleSell

Claude Code is Anthropic's official terminal-based coding agent. It runs locally, reads and edits files, executes shell commands, and uses skills, hooks, and agents you place in ~/.claude/ to customize its behavior.

Browse 220 Claude Code rules →Claude Code homepage →

Top 5 Claude Code rules right now

Ranked by quality score — freshness, schema completeness, and review signals. Refreshed daily.

Continue (continuedev): Source-Controlled AI Coding — 33K Stars

A86

@continuedev★ 33.1kClaimable

Open-source Continue: VS Code & JetBrains AI assistant with source-controlled rules, CI-enforceable checks, and a Continue CLI. Works with any provider.

UniversalAgentstodayFree

Agent Team: Content Production Pipeline

A86

@rulesell

A three-agent content pipeline — a Writer that drafts, an Editor that rewrites for clarity and impact, and an SEO Optimizer that handles meta tags, keywords, and structure. For blog posts, documentation, and marketing pages.

claude-codeagents

ClaudeAgent Teams3w agoFree

Claude Code Docs

A86

@costiashClaimable

Claude Code Plugin instant docs search, interactive HTML courses, and changelog reports. 589 official docs always up-to-date. Auto-discovery Skill, Obsidian & Amber themed courses with animated protocol diagrams, quizzes, and code translations. Zero dependencies.

ClaudeSkills3w agoFree

Claude Code System Prompts

A86

@Piebald-AI★ 9.6kClaimable

All parts of Claude Code's system prompt, 24 builtin tool descriptions, sub agent prompts (Plan/Explore/Task), utility prompts (CLAUDE.md, compact, statusline, magic docs, WebFetch, Bash cmd, security review, agent creation). Updated for each Claude Code version.

ClaudeSkills2w agoFree

MCP Server: PostgreSQL Explorer

A86

@rulesell

A Model Context Protocol server that connects Claude Code to a PostgreSQL database. Exposes schema introspection, safe read queries, explain plans, and index suggestions as MCP tools. Write operations require explicit confirmation.

mcppostgres

ClaudeMCP Servers3w agoFree

Why use Claude Code rules from RuleSell

Every Claude Code skill is quality-scored from automated signals — not star-rating gaming.
Real GitHub author attribution with SPDX license badges (MIT, Apache-2.0, etc).
Copy-paste install for ~/.claude/skills, agents, hooks, and plugins.
Claimable by the upstream author — when they claim, royalties route to them.

How to use Claude Code rules

1. Pick a Claude Code rule
Browse the listings below and open one that matches your stack.
2. Copy the file contents
Click "Copy" on the listing — or download the raw file from the source URL.
3. Paste into ~/.claude/
Drop the rule into ~/.claude/. The exact command: cp rule.md ~/.claude/agents/ # or skills/, plugins/, hooks/
4. Restart Claude Code
Restart your Claude Code session so the new rule is loaded. Verify it appears in the active context.

cp rule.md ~/.claude/agents/  # or skills/, plugins/, hooks/

Claude Code rules — frequently asked

Where do Claude Code skills live on disk?

By default in ~/.claude/skills/, ~/.claude/agents/, ~/.claude/hooks/, and ~/.claude/plugins/ — depending on the asset type. Project-scoped skills go in .claude/ inside the repo root.

How do I install a Claude Code skill from RuleSell?

Open the listing, copy the file contents, paste into the matching folder under ~/.claude/, then restart your Claude Code session so the skill is picked up.

How do I share Claude Code skills with my team?

Commit a .claude/ directory at the root of your repo with the skills, agents, and hooks you want shared. Every teammate's Claude Code session will load them automatically when they open the project.

Are RuleSell's Claude Code skills free?

All listings are free to download during beta. Paid publishing with Stripe Connect payouts opens after beta — original authors keep the majority cut.

What's the difference between a skill, an agent, and a hook?

A skill is reusable knowledge Claude can invoke during a task. An agent is a specialized sub-process Claude can dispatch with its own system prompt. A hook runs deterministically before/after specific tool calls (e.g. pre-commit validation).