Skip to content

awesome.md

Latest commit

 

History

History
228 lines (154 loc) · 13.8 KB

awesome.md

File metadata and controls

228 lines (154 loc) · 13.8 KB
Raw

Awesome Claude Code

A curated list of resources, tools, plugins, and guides for Claude Code.

Inspired by the awesome list tradition.

Contents

Official Resources

Official Plugins and Marketplace

  • Claude Code Marketplace - Official plugin marketplace for discovering and distributing Claude Code plugins.
  • Code Review Plugin - Managed code review service with multi-agent review, REVIEW.md configuration, and trigger modes. See code-review docs.
  • Code Intelligence - TypeScript - LSP-powered TypeScript/JavaScript code intelligence plugin.
  • Code Intelligence - Python - LSP-powered Python code intelligence plugin.
  • Code Intelligence - Go - LSP-powered Go code intelligence plugin.
  • Code Intelligence - Rust - LSP-powered Rust code intelligence plugin.

Find and install plugins with /plugins inside Claude Code or browse via the discover-plugins docs.

Community Plugins

Community plugins extend Claude Code for specific frameworks, tools, and workflows. Categories to look for:

  • Formatting plugins - Auto-run Prettier, Black, gofmt, rustfmt, or other formatters on file save or before commits.
  • Linting plugins - Integrate ESLint, Ruff, golangci-lint, Clippy, or other linters directly into the Claude Code workflow.
  • Framework-specific plugins - Rails helpers, Next.js generators, Django scaffolding, Laravel artisan wrappers, and similar.
  • Deployment plugins - Wrappers around Terraform, Pulumi, Kubernetes, or cloud CLIs that expose deployment actions as Claude Code skills.
  • Testing plugins - Test runner integrations (Jest, pytest, RSpec, Go test) with structured output and failure summarization.
  • Documentation plugins - Auto-generate or update inline docs, README files, or changelogs.

Browse the marketplace or search GitHub for claude-code-plugin to find community contributions.

MCP Servers for Claude Code

MCP (Model Context Protocol) servers expose external tools and data sources to Claude Code. Configure them in .mcp.json. See the MCP docs and the full MCP server directory for more.

Official and Widely Used

  • GitHub MCP - Read and write GitHub issues, pull requests, and code via the GitHub API.
  • Filesystem MCP - Expose local directories with configurable read/write access.
  • PostgreSQL MCP - Read-only query access to PostgreSQL databases.
  • Slack MCP - Post messages, read channels, and interact with Slack workspaces.
  • Playwright/Browser MCP - Browser automation and web scraping via Playwright.
  • Notion MCP - Read and write Notion pages and databases.
  • Linear MCP - Manage Linear issues, projects, and cycles.
  • Memory MCP - Persistent key-value memory that survives across sessions.
  • Fetch MCP - Fetch and parse web pages or API responses.
  • SQLite MCP - Query and inspect SQLite databases.

Templates for Custom MCP Servers

This repository includes ready-to-use MCP configuration templates:

CLAUDE.md Examples and Starters

CLAUDE.md files give Claude Code persistent, project-specific instructions. Good CLAUDE.md files dramatically improve consistency and reduce repetitive prompting.

This Repository

  • CLAUDE.md - The CLAUDE.md for this repo. Shows document index conventions, writing rules, and workflow automation notes.

Framework Starters (from templates/)

Template Stack
CLAUDE-MD-PYTHON.md Python with pytest, ruff, uv/pip/poetry, Django/FastAPI variants
CLAUDE-MD-TYPESCRIPT.md TypeScript/Node.js with ESLint, Prettier, npm/pnpm/bun
CLAUDE-MD-REACT.md React/Next.js with App Router, Tailwind, Vite, shadcn
CLAUDE-MD-RAILS.md Ruby on Rails with RSpec, Rubocop, Hotwire, API-only variant
CLAUDE-MD-MONOREPO.md Monorepo with path-specific rules, workspace conventions, imports

Tips for Writing Effective CLAUDE.md Files

  • Define your tech stack, test runner, formatter, and linter at the top.
  • Add path-specific rules using # CLAUDE.md at src/ imports for large repos.
  • Use the memory section to record architectural decisions that Claude should not override.
  • Keep rules concrete and actionable ("always run ruff check before committing") rather than vague ("write clean code").
  • See the memory docs for full CLAUDE.md syntax and features.

Guides and Tutorials

Official Docs (in this repository)

  • Best Practices - Verification strategies, prompt techniques, CLAUDE.md tips, multi-session patterns.
  • Common Workflows - Explore-fix, refactor, test, PR creation, plan mode, session management.
  • How Claude Code Works - Architecture deep dive: agentic loop, context management, tool execution.
  • Hooks Guide - Practical recipes: auto-format, file protection, notifications, verification pipelines.
  • Agent Teams - Multi-agent coordination, parallel exploration, task dependencies, quality gates.

Cheatsheets (in this repository)

The cheatsheets/ folder has 64 scannable quick-reference sheets, one per doc page. Useful for finding answers fast without reading full documentation.

Templates (in this repository)

The templates/ folder has 33 copy-paste-ready templates across skills, subagents, hooks, CI/CD, MCP, plugins, cloud providers, agent teams, and enterprise configuration.

Key starting points:

  • SKILL-TEMPLATE.md - Reference, task, research, codegen, and script-powered skill patterns.
  • SUBAGENT-TEMPLATE.md - Reviewer, debugger, domain expert, hooks, memory, background, and worktree subagent patterns.
  • HOOKS-TEMPLATE.md - Formatter, protector, notifier, auditor, verifier, and all hook event types.

Videos and Talks

  • Claude Code Announcement - See the Anthropic Blog for the original Claude Code launch post and demonstration videos.
  • Anthropic YouTube Channel - Official videos on Claude Code features and workflows.

This section will grow as the community produces tutorials, conference talks, and walkthroughs. If you have a high-quality video to suggest, see Contributing.

Tools and Utilities

Core CLI

  • claude - The main Claude Code CLI. Install with npm install -g @anthropic-ai/claude-code. See setup docs.
  • claude (Desktop) - Native desktop app with system tray integration. See desktop docs.

Companion CLI Tools

  • gh - GitHub's official CLI. Used by the GitHub MCP server and CI/CD templates for PR and issue automation.
  • jq - Lightweight JSON processor. Essential for writing hook scripts that parse or transform Claude Code's JSON output.
  • direnv - Automatically load/unload environment variables per directory. Useful for managing ANTHROPIC_API_KEY and cloud provider credentials across projects.

Debugging and Development

  • MCP Inspector - Visual tool for testing and debugging MCP servers before connecting them to Claude Code.
  • Claude Code DevContainer - Pre-configured development container with firewall rules, sandbox settings, and Codespaces support.

Cloud Provider Setup

Use Claude Code through your own cloud provider account instead of Anthropic's API directly.

Provider Template Docs
AWS Bedrock AWS-BEDROCK-SETUP.md amazon-bedrock.md
Google Vertex AI GCP-VERTEX-SETUP.md google-vertex-ai.md
Microsoft Foundry (see docs) microsoft-foundry.md
LLM Gateway/Proxy LLM-GATEWAY-SETUP.md llm-gateway.md

The LLM Gateway template covers LiteLLM-based setups with key rotation, multi-provider routing, and audit logging.

Enterprise Resources

Configuration Templates

  • ENTERPRISE-NETWORK.md - Network configuration: proxy setup, custom CA certificates, mTLS, credential refresh, and firewall rules.
  • ENTERPRISE-MANAGED-SETTINGS.md - Org-wide managed settings: permission enforcement, model restrictions, audit logging.
  • PERMISSIONS-CICD.md - CI/CD safe permission presets: read-only, test-running, controlled edits, cost controls.
  • SANDBOX-DEVCONTAINER.md - DevContainer with built-in firewall, sandbox settings, and GitHub Codespaces support.

Key Docs

Contributing

Suggestions for additions to this list are welcome. To propose a resource:

  1. Open an issue on github.com/CodeWithBehnam/cc-docs describing the resource.
  2. Include the URL, a one-line description, and which section it belongs in.
  3. For plugins and MCP servers, please verify the resource is actively maintained before suggesting.

Guidelines:

  • Only suggest real, publicly accessible resources.
  • Prefer resources with clear documentation and an open-source license.
  • Keep descriptions to one line.
  • Do not suggest resources that duplicate an already-listed item without a clear reason.

See CONTRIBUTING.md for the full contribution guide.