Install as a Claude Code plugin to gain pre-built specialist agents for code review, build-error fixing, and security scanning.
Add language-specific coding rules for TypeScript, Python, or Go so your AI assistant follows your team's conventions automatically.
Use pre-built workflow skills to run structured test-driven development steps inside your AI coding tool.
Set up automated hooks that scan for hardcoded secrets before any file is saved by the AI.
Language-specific rules must be copied manually, the plugin system cannot distribute them automatically.
ECC, short for Everything Claude Code, is a collection of configuration files, workflow definitions, and automation scripts designed to make AI coding assistants work better. It was originally built for Claude Code but has since expanded to support Cursor, OpenCode, Codex, Zed, GitHub Copilot, and other tools in the same category. The core idea is that raw AI coding tools need a layer of structure on top: rules about how to write code, specialized sub-agents that handle specific tasks, automated hooks that run checks when files change, and pre-built workflow templates called skills. The repository ships 63 specialized agents, 249 skills, and 79 legacy command shims. An agent is a named AI persona with a specific role and a limited set of tools, for example a code reviewer that only reads files, or a build error resolver that tries to fix compilation failures. A skill is a workflow template you invoke to do something structured, like running test-driven development steps or generating a security audit. Hooks are scripts that fire automatically when the AI tool takes certain actions, such as checking for hardcoded secrets before a file is saved. You install it either as a plugin directly inside Claude Code or by copying the relevant folders manually. The plugin path is the simpler route for most users. There are also rules files organized by programming language covering TypeScript, Python, Go, Swift, PHP, and others, which tell the AI assistant what conventions to follow in each language. Because the plugin system cannot distribute rules automatically, those need to be copied by hand. The project also includes a bundled security scanner called AgentShield, a minimal sandboxed AI agent runtime called MiniClaw, a desktop dashboard for browsing available components, and a cross-platform installer that works on Windows, macOS, and Linux. It is released under the MIT license and is described as a community resource where contributions of new agents, skills, and rules are encouraged. The full README is longer than what was shown.
← affaan-m on gitmyhub — every repo by this author, as a profile.
Verify against the repo before relying on details.