Everything Claude Code (ECC) - Configuration Framework & Toolkit
The ultimate configuration framework for Claude Code with 28 subagents, 119 skills, 60 commands, and AgentShield security scanner. 168K GitHub stars.
What is Everything Claude Code?#
Everything Claude Code (ECC) is a comprehensive configuration framework and toolkit for AI coding agents, primarily targeting Claude Code. Created by Anthropic hackathon winner Affaan Mustafa, ECC has become one of the most-starred repositories on GitHub with over 168,000 stars and 26,000 forks. It provides a structured, security-first approach to configuring and extending Claude Code for professional development workflows.
ECC goes far beyond a simple awesome list or collection of prompts. It is a full configuration framework with selective installation, cross-platform support (Claude Code, Cursor, Codex, OpenCode), and built-in security enforcement.
Core Components#
28 Specialized Subagents#
ECC includes 28 purpose-built subagents that handle specific development tasks:
- Planning Agent: Breaks down complex features into actionable steps
- TDD Agent: Drives test-driven development workflows
- Security Review Agent: Scans code for vulnerabilities and compliance issues
- Build Error Agent: Diagnoses and fixes compilation and build errors
- Language-Specific Code Review Agents: Separate agents for TypeScript, Python, Go, Rust, Java, Kotlin, and C++ with language-aware patterns and best practices
119 Skills#
Reusable capability modules that can be selectively installed based on your project needs. Skills range from code generation patterns to deployment automation to documentation generation.
60 Slash Commands#
Quick-access commands for common workflows. Slash commands integrate with the subagent system and can be chained together for complex multi-step operations.
AgentShield Security Scanner#
A comprehensive security layer with 1,282 tests and 102 static analysis rules. AgentShield hooks into pre-tool-use events to block dangerous git flags, detect secrets in prompts, and prevent configuration tampering. The --opus mode runs an adversarial red-team/blue-team pipeline using Claude Opus for maximum security coverage.
Key Features#
Selective Install Pipeline#
ECC uses a manifest-driven installation system backed by SQLite state tracking. Install only the languages and components you need instead of pulling the entire repository. This keeps your configuration lean and avoids conflicts.
Memory Persistence & Continuous Learning#
Session lifecycle hooks extract patterns into "instincts" with confidence scoring. Over time, your ECC configuration evolves from basic rules into full skills based on your actual development patterns. This is not just configuration — it is a learning system.
Cross-Platform Configs#
A single ECC repository generates configurations for four AI coding agents: Claude Code, Cursor, Codex, and OpenCode. The root AGENTS.md file is read by all four tools, while tool-specific configurations are generated from the same source of truth.
MCP Configurations#
Pre-configured Model Context Protocol (MCP) server setups for common integrations. These let your AI coding agent connect to external tools and data sources through standardized interfaces.
Installation#
ECC supports selective installation via its manifest-driven pipeline. Clone the repository and run the installer, selecting only the components relevant to your stack. The SQLite-backed state tracker ensures clean upgrades and prevents duplicate configurations.
Key Takeaways#
- 168K+ GitHub stars, 26K forks, 113+ contributors
- 28 subagents, 119 skills, 60 slash commands
- AgentShield with 1,282 tests and 102 static analysis rules
- Cross-platform: Claude Code, Cursor, Codex, OpenCode
- Selective install with manifest-driven pipeline
- Memory persistence and continuous learning system
- Current version: v1.9.0 (March 2026)