Claude Code Plugin Marketplace
What is this?
Claude Code is Anthropic's CLI tool for coding with AI. Plugins extend what Claude Code can do by adding specialized agents, skills, and automation hooks.
This marketplace gives you 53 ready-to-use plugins that turn Claude Code into an autonomous development powerhouse. Instead of telling Claude what to do step by step, you delegate entire tasks to specialized agents that know exactly how to handle them.
Before plugins:
You: "Read the auth module, check for SQL injection, review the error handling,
look at the input validation, check the session management..."
After plugins:
You: "Run a security review on the auth module"
Claude: *security-analyst agent handles everything autonomously*
Quick Start
Prerequisites: Claude Code and Git installed.
macOS / Linux:
git clone https://github.com/a-ariff/ariff-claude-plugins.git
cd ariff-claude-plugins && bash scripts/install.sh
Windows (PowerShell):
git clone https://github.com/a-ariff/ariff-claude-plugins.git
cd ariff-claude-plugins; .\scripts\install.ps1
Done. Start a new Claude Code session and all 65 plugins are available.
What's Inside
65 Plugins in 4 Categories
ariff-claude-plugins/
|
|-- 24 Agents autonomous workers that handle entire tasks
|-- 34 Skills methodologies and knowledge you can apply on demand
|-- 5 Hooks automatic safeguards that run without you asking
|-- 2 Commands slash commands for common workflows
|
|-- includes 12-plugin Anti-Hallucination Suite
Agents (22) -- delegate entire tasks
Agents are autonomous workers. You give them a task, they figure out the approach, use tools, read files, and deliver results.
| Category | Agents | What they do |
|---|
| Development | architect, backend-dev, frontend-dev, performance-engineer, qa-engineer, security-analyst, refactorer, autonomous-dev-assistant | Build, optimize, test, and secure code |
| Analysis | sequential-thinker, project-planner, analyzer, mentor | Plan, analyze, reason through problems |
| Safety | context-validator, intent-clarifier, pre-action-verifier, scope-boundary-checker, fact-checker, dependency-validator, rollback-planner | Validate before acting, catch mistakes early |
| Specialized | smithery-deployment-agent, smart-routing, setup-orchestrator | Deploy MCP servers, route to optimal models, configure environments |
Skills (26) -- apply proven methodologies
Skills are knowledge and methodology files that Claude loads when relevant. They teach Claude how to approach specific types of problems.
Superpowers (9): brainstorming, writing-plans, executing-plans, subagent-driven-development, verification-before-completion, root-cause-tracing, defense-in-depth, when-stuck, using-superpowers
Development (12): github, plugin-creator, plugin-checker, canvas-api, api-docs-generator, ci-cd-helper, code-reviewer, commit-message-generator, error-explainer, performance-optimizer, pr-analyzer, refactor-assistant
Testing (3): security-scanner, tdd-workflow, test-writer
Debugging (2): systematic-debugger, memory-sync
Hooks (3) -- automatic safeguards
Hooks run automatically at specific points during Claude's workflow. You don't invoke them -- they watch and act.
| Hook | Event | What it does |
|---|
| assumption-checker | PreToolUse | Warns when Claude makes assumptions instead of checking |
| memory-sync | Stop | Prompts to save learnings at session end |
| hooks-reference | PreToolUse | Reference implementation for building your own hooks |
Anti-Hallucination Suite (12 plugins) -- stop AI from guessing
