everything-claude-code-harness

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Everything Claude Code (ECC) — Agent Harness Performance System

Everything Claude Code（ECC）——Agent性能增强系统

Skill by ara.so — Daily 2026 Skills collection.

Everything Claude Code (ECC) is a production-ready performance optimization system for AI agent harnesses. It provides specialized subagents, reusable skills, custom slash commands, memory-persisting hooks, security scanning, and language-specific rules — all evolved from 10+ months of daily real-world use. Works across Claude Code, Cursor, Codex, OpenCode, and Antigravity.

由ara.so开发的Skill——属于Daily 2026 Skills合集。

Everything Claude Code（ECC）是一款可用于生产环境的AI Agent性能优化系统，它提供专用子Agent、可复用技能、自定义斜杠命令、内存持久化钩子、安全扫描以及特定语言规则——所有功能均源自10个多月的日常实际使用经验。支持Claude Code、Cursor、Codex、OpenCode和Antigravity等多个平台。

Installation

安装

Option 1: Plugin Marketplace (Recommended)

选项1：插件市场（推荐）

bash

undefined

bash

undefined

Inside Claude Code, run:

在Claude Code中运行：

/plugin marketplace add affaan-m/everything-claude-code /plugin install everything-claude-code@everything-claude-code

undefined

/plugin marketplace add affaan-m/everything-claude-code /plugin install everything-claude-code@everything-claude-code

undefined

Option 2: Manual Clone

选项2：手动克隆

bash

git clone https://github.com/affaan-m/everything-claude-code.git
cd everything-claude-code

bash

git clone https://github.com/affaan-m/everything-claude-code.git
cd everything-claude-code

Install rules for your language stack

为你的技术栈安装对应规则

./install.sh typescript

Multiple languages:

安装多语言规则：

./install.sh typescript python golang swift

Target a specific IDE:

针对特定IDE安装：

./install.sh --target cursor typescript

undefined

./install.sh --target cursor typescript

undefined

Install Rules (Always Required)

安装规则（必须执行）

Claude Code plugins cannot auto-distribute rules — install them manually via

./install.sh

or copy from

rules/

into your project's

.claude/rules/

directory.

Claude Code插件无法自动分发规则——请通过

./install.sh

手动安装，或从

rules/

目录复制到项目的

.claude/rules/

目录中。

Directory Structure

目录结构

everything-claude-code/
├── .claude-plugin/         # Plugin and marketplace manifests
│   ├── plugin.json
│   └── marketplace.json
├── agents/                 # Specialized subagents (planner, architect, etc.)
├── commands/               # Slash commands (/plan, /security-scan, etc.)
├── skills/                 # Reusable skill modules
├── hooks/                  # Lifecycle hooks (SessionStart, Stop, PostEdit, etc.)
├── rules/
│   ├── common/             # Language-agnostic rules
│   ├── typescript/
│   ├── python/
│   ├── golang/
│   └── swift/
├── scripts/                # Setup and utility scripts
└── install.sh              # Interactive installer

everything-claude-code/
├── .claude-plugin/         # 插件及市场清单
│   ├── plugin.json
│   └── marketplace.json
├── agents/                 # 专用子Agent（规划师、架构师等）
├── commands/               # 斜杠命令（/plan、/security-scan等）
├── skills/                 # 可复用技能模块
├── hooks/                  # 生命周期钩子（SessionStart、Stop、PostEdit等）
├── rules/
│   ├── common/             # 通用语言无关规则
│   ├── typescript/
│   ├── python/
│   ├── golang/
│   └── swift/
├── scripts/                # 安装与实用脚本
└── install.sh              # 交互式安装程序

Key Commands

关键命令

After installation, use the namespaced form (plugin install) or short form (manual install):

bash

undefined

安装完成后，可使用命名空间形式（插件安装）或简写形式（手动安装）：

bash

undefined

Planning & architecture

规划与架构

/everything-claude-code:plan "Add OAuth2 login flow" /everything-claude-code:architect "Design a multi-tenant SaaS system"

/everything-claude-code:plan "添加OAuth2登录流程" /everything-claude-code:architect "设计多租户SaaS系统"

Research-first development

先调研再开发

/everything-claude-code:research "Best approach for rate limiting in Node.js"

/everything-claude-code:research "Node.js中限流的最佳方案"

Security

安全相关

/everything-claude-code:security-scan /everything-claude-code:harness-audit

Agent loops and orchestration

Agent循环与编排

/everything-claude-code:loop-start /everything-claude-code:loop-status /everything-claude-code:quality-gate /everything-claude-code:model-route

Multi-agent workflows

多Agent工作流

/everything-claude-code:multi-plan /everything-claude-code:multi-execute /everything-claude-code:multi-backend /everything-claude-code:multi-frontend

Session and memory

会话与内存

/everything-claude-code:sessions /everything-claude-code:instinct-import

PM2 orchestration

PM2编排

/everything-claude-code:pm2

Package manager setup

包管理器设置

/everything-claude-code:setup-pm


> With manual install, drop the `everything-claude-code:` prefix: `/plan`, `/sessions`, etc.

---

/everything-claude-code:setup-pm


> 手动安装时，可省略`everything-claude-code:`前缀：直接使用`/plan`、`/sessions`等。

---

Hook Runtime Controls

Hook运行时控制

ECC hooks fire at agent lifecycle events. Control strictness at runtime without editing files:

bash

undefined

ECC钩子会在Agent生命周期事件触发时执行。无需修改文件即可在运行时控制严格程度：

bash

undefined

Set hook strictness profile

设置钩子严格度配置

export ECC_HOOK_PROFILE=minimal # Least intrusive export ECC_HOOK_PROFILE=standard # Default export ECC_HOOK_PROFILE=strict # Maximum enforcement

export ECC_HOOK_PROFILE=minimal # 最低侵入性 export ECC_HOOK_PROFILE=standard # 默认配置 export ECC_HOOK_PROFILE=strict # 最高强制级别

Disable specific hooks by ID (comma-separated)

通过ID禁用特定钩子（逗号分隔）

export ECC_DISABLED_HOOKS="pre:bash:tmux-reminder,post:edit:typecheck"


Hook events covered: `SessionStart`, `Stop`, `PostEdit`, `PreBash`, `PostBash`, and more.

---

export ECC_DISABLED_HOOKS="pre:bash:tmux-reminder,post:edit:typecheck"


支持的钩子事件包括：`SessionStart`、`Stop`、`PostEdit`、`PreBash`、`PostBash`等。

---

Package Manager Detection

包管理器自动检测

ECC auto-detects your package manager with this priority chain:

```
CLAUDE_PACKAGE_MANAGER
```
environment variable
```
.claude/package-manager.json
```
(project-level)
```
package.json
```
→
```
packageManager
```
field

Lock file detection (

package-lock.json

yarn.lock

pnpm-lock.yaml

bun.lockb

)

```
~/.claude/package-manager.json
```
(global)
First available manager as fallback

bash

undefined

ECC会按以下优先级自动检测你的包管理器：

```
CLAUDE_PACKAGE_MANAGER
```
环境变量
```
.claude/package-manager.json
```
（项目级）
```
package.json
```
中的
```
packageManager
```
字段

锁定文件检测（

package-lock.json

、

yarn.lock

、

pnpm-lock.yaml

、

bun.lockb

）

```
~/.claude/package-manager.json
```
（全局）
回退到第一个可用的包管理器

bash

undefined

Set via environment

通过环境变量设置

export CLAUDE_PACKAGE_MANAGER=pnpm

Set globally

全局设置

node scripts/setup-package-manager.js --global pnpm

Set per-project

项目级设置

node scripts/setup-package-manager.js --project bun

Detect current setting

检测当前设置

node scripts/setup-package-manager.js --detect

---

node scripts/setup-package-manager.js --detect

---

Skills System

技能系统

Skills are markdown modules the agent loads to gain domain expertise. Install individually or in bulk.

技能是Agent加载以获取领域专业知识的Markdown模块，可单独或批量安装。

Using a Skill

使用技能

bash

undefined

bash

undefined

Reference a skill explicitly in your prompt

在提示词中明确引用技能

"Use the search-first skill to find the right caching approach before implementing"

"使用search-first技能在实现前找到合适的缓存方案"

Or trigger via slash command

或通过斜杠命令触发

/everything-claude-code:research "content hashing strategies for API responses"

undefined

/everything-claude-code:research "API响应的内容哈希策略"

undefined

Notable Built-in Skills

内置重要技能

Skill	Purpose
`search-first`	Research before coding — avoids hallucinated APIs
`cost-aware-llm-pipeline`	Optimizes token spend across model calls
`content-hash-cache-pattern`	Cache invalidation via content hashing
`skill-stocktake`	Audits which skills are loaded and active
`frontend-slides`	Zero-dependency HTML presentation builder
`configure-ecc`	Guided interactive ECC setup wizard
`swift-actor-persistence`	Swift concurrency + persistence patterns
`regex-vs-llm-structured-text`	Decides when to use regex vs LLM parsing

技能	用途
`search-first`	编码前先调研——避免生成不存在的API
`cost-aware-llm-pipeline`	优化跨模型调用的Token消耗
`content-hash-cache-pattern`	通过内容哈希实现缓存失效
`skill-stocktake`	审计已加载并激活的技能
`frontend-slides`	零依赖HTML演示文稿构建工具
`configure-ecc`	交互式ECC引导设置向导
`swift-actor-persistence`	Swift并发+持久化模式
`regex-vs-llm-structured-text`	决定何时使用正则表达式而非LLM解析

Writing a Custom Skill

编写自定义技能

Create

skills/my-skill.md

markdown

---
name: my-skill
description: What this skill does
triggers:
  - "phrase that activates this skill"
---

创建

skills/my-skill.md

：

markdown

---
name: my-skill
description: 该技能的用途
triggers:
  - "激活该技能的短语"
---

My Skill

我的技能

When to Use

使用场景

...

Pattern

模式

```typescript // concrete example ```

```typescript // 具体示例 ```

Rules

规则

Rule one
Rule two

---

规则一
规则二

---

Instincts System (Continuous Learning)

直觉系统（持续学习）

Instincts are session-extracted patterns saved for reuse. They carry confidence scores and evolve over time.

直觉是从会话中提取的可复用模式，带有置信度评分并会随时间演进。

Export an Instinct

导出直觉

bash

/everything-claude-code:instinct-import

bash

/everything-claude-code:instinct-import

Instinct File Format

直觉文件格式

markdown

---
name: prefer-zod-for-validation
confidence: 0.92
extracted_from: session-2026-02-14
---

markdown

---
name: prefer-zod-for-validation
confidence: 0.92
extracted_from: session-2026-02-14
---

Action

操作建议

Always use Zod for runtime schema validation in TypeScript projects.

在TypeScript项目中始终使用Zod进行运行时 schema 验证。

Evidence

依据

Caught 3 runtime type errors that TypeScript alone missed during session.

在会话中发现3个仅靠TypeScript无法检测到的运行时类型错误。

Examples

示例

```typescript import { z } from 'zod'

const UserSchema = z.object({ id: z.string().uuid(), email: z.string().email(), role: z.enum(['admin', 'user']) })

type User = z.infer<typeof UserSchema> ```

---

```typescript import { z } from 'zod'

const UserSchema = z.object({ id: z.string().uuid(), email: z.string().email(), role: z.enum(['admin', 'user']) })

type User = z.infer<typeof UserSchema> ```

---

Rules Architecture

规则架构

Rules enforce coding standards per language. Install only what your stack needs.

bash

undefined

规则用于按语言强制执行编码标准，仅安装你的技术栈所需的规则即可。

bash

undefined

TypeScript + Python

安装TypeScript + Python规则

./install.sh typescript python

Check what's installed

查看已安装规则

ls .claude/rules/

undefined

ls .claude/rules/

undefined

Rule Directory Layout

规则目录结构

rules/
├── common/         # Applies to all languages
│   ├── research-first.md
│   ├── security-baseline.md
│   └── verification-loops.md
├── typescript/
│   ├── no-any.md
│   ├── zod-validation.md
│   └── strict-mode.md
├── python/
│   ├── type-hints.md
│   └── django-patterns.md
└── golang/
    └── error-wrapping.md

rules/
├── common/         # 适用于所有语言
│   ├── research-first.md
│   ├── security-baseline.md
│   └── verification-loops.md
├── typescript/
│   ├── no-any.md
│   ├── zod-validation.md
│   └── strict-mode.md
├── python/
│   ├── type-hints.md
│   └── django-patterns.md
└── golang/
    └── error-wrapping.md

Agents (Subagent Delegation)

Agent（子Agent委托）

Agents are specialized personas the orchestrator delegates to:

bash

undefined

Agent是编排器可委托的专用角色：

bash

undefined

In your prompt, reference an agent explicitly

在提示词中明确引用Agent

"Delegate architecture decisions to the architect agent" "Use the planner agent to break this feature into tasks"


Available agents include: `planner`, `architect`, `researcher`, `verifier`, `security-auditor`, and more. Each lives in `agents/<name>.md` with its own system prompt, tools list, and constraints.

---

"将架构决策委托给architect agent" "使用planner agent将该功能拆分为任务"


可用Agent包括：`planner`、`architect`、`researcher`、`verifier`、`security-auditor`等。每个Agent都存放在`agents/<name>.md`中，包含各自的系统提示词、工具列表与约束条件。

---

AgentShield Security Scanning

AgentShield安全扫描

Run security scans directly from Claude Code:

bash

/everything-claude-code:security-scan

This invokes the AgentShield scanner (1282 tests, 102 rules) against your codebase and surfaces:

Hardcoded secrets
Injection vulnerabilities
Insecure dependencies
Agent prompt injection patterns

直接在Claude Code中运行安全扫描：

bash

/everything-claude-code:security-scan

这会调用AgentShield扫描器（包含1282项测试、102条规则）扫描你的代码库，并识别以下问题：

硬编码密钥
注入漏洞
不安全依赖
Agent提示词注入模式

Memory Persistence Hooks

内存持久化钩子

ECC hooks automatically save and restore session context:

javascript

// hooks/session-start.js — loads prior context on new session
const fs = require('fs')
const path = require('path')

const memoryPath = path.join(process.env.HOME, '.claude', 'session-memory.json')

if (fs.existsSync(memoryPath)) {
  const memory = JSON.parse(fs.readFileSync(memoryPath, 'utf8'))
  console.log('Restored session context:', memory.summary)
}

javascript

// hooks/stop.js — saves session summary on exit
const summary = {
  timestamp: new Date().toISOString(),
  summary: process.env.ECC_SESSION_SUMMARY || '',
  skills_used: (process.env.ECC_SKILLS_USED || '').split(',')
}

fs.writeFileSync(memoryPath, JSON.stringify(summary, null, 2))

ECC钩子会自动保存与恢复会话上下文：

javascript

// hooks/session-start.js —— 在新会话加载之前的上下文
const fs = require('fs')
const path = require('path')

const memoryPath = path.join(process.env.HOME, '.claude', 'session-memory.json')

if (fs.existsSync(memoryPath)) {
  const memory = JSON.parse(fs.readFileSync(memoryPath, 'utf8'))
  console.log('Restored session context:', memory.summary)
}

javascript

// hooks/stop.js —— 在退出时保存会话摘要
const summary = {
  timestamp: new Date().toISOString(),
  summary: process.env.ECC_SESSION_SUMMARY || '',
  skills_used: (process.env.ECC_SKILLS_USED || '').split(',')
}

fs.writeFileSync(memoryPath, JSON.stringify(summary, null, 2))

Cross-Platform Support

跨平台支持

Platform	Support
Claude Code	Full (agents, commands, skills, hooks, rules)
Cursor	Full (via `--target cursor` installer flag)
OpenCode	Full (plugin system, 20+ hook event types, 3 native tools)
Codex CLI	Full ( `codex.md` generated via `/codex-setup` )
Codex App	Full ( `AGENTS.md` -based)
Antigravity	Full (via `--target antigravity` installer flag)

平台	支持情况
Claude Code	完全支持（Agent、命令、技能、钩子、规则）
Cursor	完全支持（通过安装器 `--target cursor` 参数）
OpenCode	完全支持（插件系统、20+钩子事件类型、3个原生工具）
Codex CLI	完全支持（通过 `/codex-setup` 生成 `codex.md` ）
Codex App	完全支持（基于 `AGENTS.md` ）
Antigravity	完全支持（通过安装器 `--target antigravity` 参数）

Common Patterns

常用模式

Research-First Development

先调研再开发

"Before implementing the payment webhook handler, use the search-first skill to 
verify current Stripe webhook verification best practices."

"在实现支付Webhook处理器之前，使用search-first技能验证当前Stripe Webhook验证的最佳实践。"

Token Optimization

Token优化

bash

undefined

bash

undefined

Route to cheaper model for simple tasks

将简单任务路由到更便宜的模型

/everything-claude-code:model-route "Write a unit test for this pure function"

/everything-claude-code:model-route "为这个纯函数编写单元测试"

Use background processes for long analysis

使用后台进程进行长时间分析

/everything-claude-code:harness-audit

undefined

/everything-claude-code:harness-audit

undefined

Parallelization with Git Worktrees

基于Git工作树的并行化

bash

undefined

bash

undefined

Create isolated worktrees for parallel agent tasks

为并行Agent任务创建独立工作树

git worktree add ../feature-auth -b feature/auth git worktree add ../feature-payments -b feature/payments

Each Claude Code session operates in its own worktree

每个Claude Code会话在各自的工作树中运行

Merge when both complete

完成后合并

undefined

undefined

Verification Loop

验证循环

bash

/everything-claude-code:loop-start    # Begin tracked loop

bash

/everything-claude-code:loop-start    # 开始跟踪循环

... agent does work ...

... Agent执行工作 ...

/everything-claude-code:loop-status # Check progress /everything-claude-code:quality-gate # Enforce pass criteria before merge

---

/everything-claude-code:loop-status # 检查进度 /everything-claude-code:quality-gate # 合并前强制执行通过标准

---

Troubleshooting

故障排查

Plugin commands not found after install

bash

/plugin list everything-claude-code@everything-claude-code

安装后找不到插件命令

bash

/plugin list everything-claude-code@everything-claude-code

If empty, re-run: /plugin install everything-claude-code@everything-claude-code

如果为空，重新运行：/plugin install everything-claude-code@everything-claude-code


**Rules not applied**
```bash


**规则未生效**
```bash

Rules require manual install — plugin system cannot distribute them

规则需要手动安装——插件系统无法分发规则

cd everything-claude-code && ./install.sh typescript

Verify:

验证：

ls ~/.claude/rules/ # or .claude/rules/ in project root


**Hooks not firing**
```bash

ls ~/.claude/rules/ # 或项目根目录下的.claude/rules/


**钩子未触发**
```bash

Check profile setting

检查配置文件设置

echo $ECC_HOOK_PROFILE

Check disabled list

检查禁用列表

echo $ECC_DISABLED_HOOKS

Reset to defaults

重置为默认值

unset ECC_HOOK_PROFILE unset ECC_DISABLED_HOOKS


**Instinct import drops content**
Ensure you're on v1.4.1+. Earlier versions had a bug where `parse_instinct_file()` silently dropped Action/Evidence/Examples sections. Pull latest and re-run.

**Wrong package manager used**
```bash
node scripts/setup-package-manager.js --detect
export CLAUDE_PACKAGE_MANAGER=pnpm   # Override explicitly

unset ECC_HOOK_PROFILE unset ECC_DISABLED_HOOKS


**直觉导入丢失内容**
确保你使用的是v1.4.1及以上版本。早期版本中`parse_instinct_file()`存在静默丢失Action/Evidence/Examples章节的Bug。拉取最新版本后重新运行。

**使用了错误的包管理器**
```bash
node scripts/setup-package-manager.js --detect
export CLAUDE_PACKAGE_MANAGER=pnpm   # 显式覆盖

Resources

资源

Homepage: https://ecc.tools
GitHub: https://github.com/affaan-m/everything-claude-code
GitHub App (Marketplace): https://github.com/marketplace/ecc-tools
npm (universal):
```
ecc-universal
```
npm (security):
```
ecc-agentshield
```
Shorthand Guide: https://x.com/affaanmustafa/status/2012378465664745795
Longform Guide: https://x.com/affaanmustafa/status/2014040193557471352

主页：https://ecc.tools
GitHub：https://github.com/affaan-m/everything-claude-code
GitHub应用（市场）：https://github.com/marketplace/ecc-tools
npm（通用）：
```
ecc-universal
```
npm（安全）：
```
ecc-agentshield
```
速查指南：https://x.com/affaanmustafa/status/2012378465664745795
详细指南：https://x.com/affaanmustafa/status/2014040193557471352