// curated GitHub index // 58 repositories // verified 4 August 2026
Find Claude Code tools you can understand before you install.
A deliberately curated index of plugins, skills, workflows, hooks, and tools. Every repository passed a 1K-star floor and recent-maintenance check, with live GitHub stars, practical use cases, and security cautions.
Start with what you want to do
Choose more than one tag to narrow the list to projects that match every selection.
Claude Code is Anthropic's coding assistant for the terminal and supported editors. It can read your project, edit files, run commands, and help with Git using natural-language requests. This repository is where Anthropic publishes installers, bundled plugins, release notes, and public issue tracking. Beginners should use the official documentation for setup, then use this repository to check what changed in a version or whether a problem is already known.
Best for
Installing Claude Code from the official source, following releases, and checking whether Anthropic already knows about a bug.
Useful for
Review details and install
Why it is interesting
It is the closest public view of what ships in Claude Code and how the product changes over time.
Check before installing
An issue or discussion is a user report, not official documentation. Confirm security-sensitive or version-specific behaviour in Anthropic's docs and release notes.
Repository-specificactive 4 August 2026source Awesome Claude CodeClaude Code
This is Anthropic's official catalog for finding Claude Code plugins. A plugin adds a focused capability, such as a command, skill, agent, hook, or connection to another service. You can browse the directory from Claude Code and install a chosen plugin without manually copying its files. The catalog includes both Anthropic-maintained and third-party projects, so directory inclusion does not remove the need to review what a plugin can run.
Best for
Beginners who want to discover an extra Claude Code capability through the supported plugin marketplace.
Useful for
Review details and install
Why it is interesting
It shows both ready-to-install plugins and the current folder and metadata structure expected by Claude Code.
Check before installing
Official directory inclusion is not a guarantee that third-party code is safe. Review each plugin's hooks, MCP servers, scripts, permissions, and source repository.
Superpowers gives your coding agent a step-by-step software-development process instead of letting it immediately start writing code. It first helps clarify the idea, turn it into a design, and break the work into a plan you can review. Its skills then guide implementation, testing, debugging, code review, and finishing the branch. This is useful when you want a repeatable engineering workflow, but it also makes the agent more structured and deliberate than a normal chat.
Best for
People who want the agent to ask questions, create a reviewed plan, test its work, and follow the same development process each time.
Useful for
Review details and install
Why it is interesting
It combines many small skills into one end-to-end methodology and supports several coding-agent runtimes.
Check before installing
It changes how the agent approaches most development tasks and can add process to small jobs. Review its startup hook and skills so you understand when they activate.
Coding agents often spend many tokens explaining what they are doing, even when a short answer would be enough. Caveman changes the agent's writing style so replies are compressed while code, paths, commands, and technical details stay intact. Shorter replies can reduce output-token usage, cost less to read back into a long conversation, and make the session easier to scan. It mainly changes how much the agent says; it does not make the underlying model smaller or guarantee the same saving on every coding task.
Best for
People who want concise agent replies, lower conversational token usage, and less text between code changes.
Useful for
Review details and install
Why it is interesting
It includes several compression levels, session activation, token-usage statistics, and tools for shortening memory files and MCP descriptions.
Check before installing
The repository's benchmark measures specific tasks and configurations, so treat its percentage savings as examples rather than a guarantee. Use a less aggressive mode when security explanations or destructive actions need more context.
Documented install
claude plugin marketplace add JuliusBrussee/cavemanclaude plugin install caveman@caveman
MITactive 4 August 2026source Awesome Claude CodeClaude CodeCodexGemini CLICursor
Ponytail teaches your coding agent to look for the simplest safe solution before it writes new code. It asks whether the feature is needed, whether the project already solves it, and whether a built-in tool or installed dependency can do the job. Because the agent writes and reviews less unnecessary code, it can save tokens, avoid extra dependencies, and leave smaller changes for you to understand. It does not mean skipping tests, security, accessibility, or error handling; those safeguards remain part of the work.
Best for
People whose coding agent tends to over-engineer simple requests, create unnecessary abstractions, or add dependencies too quickly.
Useful for
Review details and install
Why it is interesting
Its decision ladder makes reuse and minimal code an always-on habit, with separate commands for reviewing a change or auditing a whole repository.
Check before installing
The Claude Code plugin uses lifecycle hooks and requires Node on the non-interactive PATH. Minimalism should not remove required validation, accessibility, or safety controls.
A single CLAUDE. Best suited to developers who want a tiny, low-friction instruction file that counters common AI coding mistakes.
Best for
developers who want a tiny, low-friction instruction file that counters common AI coding mistakes.
Useful for
Review details and install
Why it is interesting
It distills four memorable coding-agent behaviors into a single CLAUDE.md file.
Check before installing
Read the SKILL.md, scripts, hooks, requested tools, and network behavior before installation. Test new skills in a disposable project before trusting them with sensitive repositories.
Repository-specificactive 20 April 2026source Awesome Claude CodeClaude Code
Public repository for Agent Skills. Best suited to learning the official Agent Skills format and starting from Anthropic-maintained examples.
Best for
learning the official Agent Skills format and starting from Anthropic-maintained examples.
Useful for
Review details and install
Why it is interesting
It is the primary public reference for the SKILL.md standard used by Claude Code.
Check before installing
Official status does not remove version, permission, or supply-chain risk. Follow the repository's current documentation and use least-privilege credentials.
Repository-specificactive 24 July 2026source Awesome Claude CodeClaude Code
Use Garry Tan's exact Claude Code setup: 23 opinionated tools that serve as CEO, Designer, Eng Manager, Release Manager, Doc Engineer, and QA. Best suited to founders and product teams that want opinionated roles for planning, design, engineering, release, and QA.
Best for
founders and product teams that want opinionated roles for planning, design, engineering, release, and QA.
Useful for
Review details and install
Why it is interesting
Its role-based tools package an end-to-end startup product workflow for Claude Code.
Check before installing
Opinionated automation can change files, Git state, and task scope. Start with a small repository and keep human approval around destructive or publishing steps.
MITactive 15 July 2026source Awesome Claude CodeClaude Code
Bash is all you need - A nano claude code–like 「agent harness」, built from 0 to 1. Best suited to developers who want to understand an agent harness by building a small one from first principles.
Best for
developers who want to understand an agent harness by building a small one from first principles.
Useful for
Review details and install
Why it is interesting
Its compact Bash implementation makes the agent loop easier to inspect than a production-scale codebase.
Check before installing
Examples can lag behind fast-moving Claude Code releases. Cross-check commands and security-sensitive guidance against current Anthropic documentation.
MITactive 28 July 2026source Awesome Claude CodeClaude Code
A visual, example-driven guide to Claude Code — from basic concepts to advanced agents, with copy-paste templates that bring immediate value. Best suited to beginners who prefer visual explanations, progressive lessons, and copy-ready templates.
Best for
beginners who prefer visual explanations, progressive lessons, and copy-ready templates.
Useful for
Review details and install
Why it is interesting
It combines a structured learning path with practical examples across Claude Code's major features.
Check before installing
Examples can lag behind fast-moving Claude Code releases. Cross-check commands and security-sensitive guidance against current Anthropic documentation.
MITactive 4 August 2026source Awesome Claude CodeClaude Code
Security scanner for AI agent skills. Best suited to scanning third-party agent skills for malicious patterns and risky behavior before installation.
Best for
scanning third-party agent skills for malicious patterns and risky behavior before installation.
Useful for
Review details and install
Why it is interesting
It applies security-focused static analysis specifically to AI agent skill packages.
Check before installing
A guardrail or scanner reduces risk but cannot prove that agent activity is safe. Keep backups, least privilege, review, and platform controls in place.
Apache-2.0active 4 August 2026source Awesome Claude CodeClaude Code
Use Claude Code, OpenCode, Cursor CLI, and Codex on mobile and web with CloudCLI (aka Claude Code UI). Best suited to using and managing Claude Code, Codex, Cursor CLI, or OpenCode from a browser or mobile device.
Best for
using and managing Claude Code, Codex, Cursor CLI, or OpenCode from a browser or mobile device.
Useful for
Review details and install
Why it is interesting
It provides a self-hostable web interface for multiple coding-agent CLIs.
Check before installing
Remote and graphical clients may expose terminals, source code, or agent approvals. Review authentication, network exposure, update behavior, and local data storage.
AGPL-3.0active 4 August 2026source Awesome Claude CodeClaude Code
Self-organizing AI second brain for Obsidian + Claude Code. Best suited to Obsidian users building a local Markdown knowledge base with Claude Code.
Best for
Obsidian users building a local Markdown knowledge base with Claude Code.
Useful for
Review details and install
Why it is interesting
It organizes imported material into an interconnected, user-owned note graph.
Check before installing
Persistent context systems may retain source code, prompts, tool output, or personal notes. Review storage, deletion, encryption, and model-provider boundaries first.
MITactive 1 August 2026source Awesome Claude CodeClaude Code
Examples of my Claude Code infrastructure with skill auto-activation, hooks, and agents. Best suited to studying a real Claude Code setup with auto-activated skills, hooks, and agents.
Best for
studying a real Claude Code setup with auto-activated skills, hooks, and agents.
Useful for
Review details and install
Why it is interesting
It exposes the configuration patterns behind a mature personal Claude Code environment.
Check before installing
Examples can lag behind fast-moving Claude Code releases. Cross-check commands and security-sensitive guidance against current Anthropic documentation.
MITactive 13 July 2026source Awesome Claude CodeClaude Code
Autonomous AI development loop for Claude Code with intelligent exit detection. Best suited to running iterative autonomous development loops with explicit completion detection.
Best for
running iterative autonomous development loops with explicit completion detection.
Useful for
Review details and install
Why it is interesting
It adapts the Ralph loop pattern to Claude Code with safeguards against endless execution.
Check before installing
Opinionated automation can change files, Git state, and task scope. Start with a small repository and keep human approval around destructive or publishing steps.
MITactive 18 July 2026source Awesome Claude CodeClaude Code
Agent skill that generates rich HTML pages or slide decks for diagrams, diff reviews, plan audits, data tables, and project recaps. Best suited to generating readable HTML diagrams, plan reviews, data tables, slide decks, and project recaps.
Best for
generating readable HTML diagrams, plan reviews, data tables, slide decks, and project recaps.
Useful for
Review details and install
Why it is interesting
It gives coding agents a polished visual output format for information that prose explains poorly.
Check before installing
Read the SKILL.md, scripts, hooks, requested tools, and network behavior before installation. Test new skills in a disposable project before trusting them with sensitive repositories.
MITactive 25 June 2026source Awesome Claude CodeClaude Code
Claude Code Action brings Claude-powered coding workflows into GitHub pull requests and issues. Best suited to adding official Claude-powered workflows to GitHub issues and pull requests.
Best for
adding official Claude-powered workflows to GitHub issues and pull requests.
Useful for
Review details and install
Why it is interesting
It is Anthropic's maintained bridge between Claude Code and GitHub Actions.
Check before installing
Official status does not remove version, permission, or supply-chain risk. Follow the repository's current documentation and use least-privilege credentials.
MITactive 4 August 2026source Awesome Claude CodeClaude Code
A Claude Skill to give your agent the ability to use a web browser. Best suited to letting Claude Code operate and verify a browser during frontend development.
Best for
letting Claude Code operate and verify a browser during frontend development.
Useful for
Review details and install
Why it is interesting
It exposes browser automation through a skill-oriented workflow designed for coding agents.
Check before installing
Read the SKILL.md, scripts, hooks, requested tools, and network behavior before installation. Test new skills in a disposable project before trusting them with sensitive repositories.
MITactive 15 July 2026source Awesome Claude CodeClaude Code
An AI-powered security review GitHub Action using Claude to analyze code changes for security vulnerabilities. Best suited to running AI-assisted security analysis on pull-request diffs in GitHub Actions.
Best for
running AI-assisted security analysis on pull-request diffs in GitHub Actions.
Useful for
Review details and install
Why it is interesting
It provides an official starting point for automated vulnerability review with Claude.
Check before installing
A guardrail or scanner reduces risk but cannot prove that agent activity is safe. Keep backups, least privilege, review, and platform controls in place.
MITactive 11 February 2026source Awesome Claude CodeClaude Code
Claude Code Guide - Setup, Commands, workflows, agents, skills & tips-n-tricks from beginner to power user! Best suited to finding a current single-page reference for setup, commands, hooks, agents, skills, and workflows.
Best for
finding a current single-page reference for setup, commands, hooks, agents, skills, and workflows.
Useful for
Review details and install
Why it is interesting
It prioritizes practical discovery and stays close to Claude Code's changing command surface.
Check before installing
Examples can lag behind fast-moving Claude Code releases. Cross-check commands and security-sensitive guidance against current Anthropic documentation.
MITactive 4 August 2026source Awesome Claude CodeClaude Code
Claude Code Dedicated Development Harness - Achieving High-Quality Development Through an Autonomous Plan→Work→Review Cycle. Best suited to teams that want a repeatable plan, implementation, and review cycle around Claude Code.
Best for
teams that want a repeatable plan, implementation, and review cycle around Claude Code.
Useful for
Review details and install
Why it is interesting
It packages quality gates around the full development loop rather than a single prompt.
Check before installing
Opinionated automation can change files, Git state, and task scope. Start with a small repository and keep human approval around destructive or publishing steps.
MITactive 4 August 2026source Awesome Claude CodeClaude Code
Skill that audits and rewrites content to remove AI writing patterns. Best suited to auditing and rewriting prose that contains repetitive or synthetic-sounding AI patterns.
Best for
auditing and rewriting prose that contains repetitive or synthetic-sounding AI patterns.
Useful for
Review details and install
Why it is interesting
It gives multiple agent runtimes a focused editorial cleanup skill.
Check before installing
Read the SKILL.md, scripts, hooks, requested tools, and network behavior before installation. Test new skills in a disposable project before trusting them with sensitive repositories.
MITactive 4 August 2026source Awesome Claude CodeClaude Code
A practical Claude Code guide with clear mental models and copy-paste examples — setup, prompt engineering, slash commands, skills, hooks, subagents, agent teams, and MCP servers. Best suited to moving from beginner concepts to advanced Claude Code workflows in one practical guide.
Best for
moving from beginner concepts to advanced Claude Code workflows in one practical guide.
Useful for
Review details and install
Why it is interesting
Its mental-model-first structure connects setup, prompting, skills, hooks, agents, teams, and MCP.
Check before installing
Examples can lag behind fast-moving Claude Code releases. Cross-check commands and security-sensitive guidance against current Anthropic documentation.
MITactive 28 July 2026source Awesome Claude CodeClaude Code
