self-improving-agent

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Self-Improving Agent

自改进代理（Self-Improving Agent）

Auto-memory captures. This plugin curates.

Claude Code's auto-memory (v2.1.32+) automatically records project patterns, debugging insights, and your preferences in

MEMORY.md

. This plugin adds the intelligence layer: it analyzes what Claude has learned, promotes proven patterns into project rules, and extracts recurring solutions into reusable skills.

自动内存负责捕获，本插件负责整理。

Claude Code的自动内存（v2.1.32+）会自动在

MEMORY.md

中记录项目模式、调试见解和您的偏好。本插件增加了智能层：它会分析Claude的学习内容，将已验证的模式推广为项目规则，并将重复出现的解决方案提取为可复用的Skill。

Quick Reference

快速参考

Command	What it does
`/si:review`	Analyze MEMORY.md — find promotion candidates, stale entries, consolidation opportunities
`/si:promote`	Graduate a pattern from MEMORY.md → CLAUDE.md or `.claude/rules/`
`/si:extract`	Turn a proven pattern into a standalone skill
`/si:status`	Memory health dashboard — line counts, topic files, recommendations
`/si:remember`	Explicitly save important knowledge to auto-memory

命令	功能
`/si:review`	分析MEMORY.md — 找出可推广候选条目、过时条目和可合并的内容
`/si:promote`	将模式从MEMORY.md升级至CLAUDE.md或 `.claude/rules/`
`/si:extract`	将已验证的模式转化为独立Skill
`/si:status`	内存健康仪表盘 — 行数统计、主题文件、优化建议
`/si:remember`	将重要知识明确保存至自动内存

How It Fits Together

协作机制

┌─────────────────────────────────────────────────────────┐
│                  Claude Code Memory Stack                │
├─────────────┬──────────────────┬────────────────────────┤
│  CLAUDE.md  │   Auto Memory    │   Session Memory       │
│  (you write)│   (Claude writes)│   (Claude writes)      │
│  Rules &    │   MEMORY.md      │   Conversation logs    │
│  standards  │   + topic files  │   + continuity         │
│  Full load  │   First 200 lines│   Contextual load      │
├─────────────┴──────────────────┴────────────────────────┤
│              ↑ /si:promote        ↑ /si:review          │
│         Self-Improving Agent (this plugin)               │
│              ↓ /si:extract    ↓ /si:remember            │
├─────────────────────────────────────────────────────────┤
│  .claude/rules/    │    New Skills    │   Error Logs     │
│  (scoped rules)    │    (extracted)   │   (auto-captured)│
└─────────────────────────────────────────────────────────┘

┌─────────────────────────────────────────────────────────┐
│                  Claude Code Memory Stack                │
├─────────────┬──────────────────┬────────────────────────┤
│  CLAUDE.md  │   Auto Memory    │   Session Memory       │
│  (你编写)│   (Claude编写)│   (Claude编写)      │
│  规则与    │   MEMORY.md      │   对话日志    │
│  标准  │   + 主题文件  │   + 会话连续性         │
│  完整加载  │   前200行│   上下文加载      │
├─────────────┴──────────────────┴────────────────────────┤
│              ↑ /si:promote        ↑ /si:review          │
│         Self-Improving Agent (本插件)               │
│              ↓ /si:extract    ↓ /si:remember            │
├─────────────────────────────────────────────────────────┤
│  .claude/rules/    │    New Skills    │   Error Logs     │
│  (范围规则)    │    (提取生成)   │   (自动捕获)│
└─────────────────────────────────────────────────────────┘

Installation

安装方法

Claude Code (Plugin)

Claude Code（插件）

/plugin marketplace add alirezarezvani/claude-skills
/plugin install self-improving-agent@claude-code-skills

/plugin marketplace add alirezarezvani/claude-skills
/plugin install self-improving-agent@claude-code-skills

OpenClaw

bash

clawhub install self-improving-agent

bash

clawhub install self-improving-agent

Codex CLI

bash

./scripts/codex-install.sh --skill self-improving-agent

bash

./scripts/codex-install.sh --skill self-improving-agent

Memory Architecture

内存架构

Where things live

文件存储位置

File	Who writes	Scope	Loaded
`./CLAUDE.md`	You (+ `/si:promote` )	Project rules	Full file, every session
`~/.claude/CLAUDE.md`	You	Global preferences	Full file, every session
`~/.claude/projects/<path>/memory/MEMORY.md`	Claude (auto)	Project learnings	First 200 lines
`~/.claude/projects/<path>/memory/*.md`	Claude (overflow)	Topic-specific notes	On demand
`.claude/rules/*.md`	You (+ `/si:promote` )	Scoped rules	When matching files open

文件	编写者	范围	加载方式
`./CLAUDE.md`	你（+ `/si:promote` ）	项目规则	每次会话完整加载
`~/.claude/CLAUDE.md`	你	全局偏好	每次会话完整加载
`~/.claude/projects/<path>/memory/MEMORY.md`	Claude（自动）	项目学习内容	加载前200行
`~/.claude/projects/<path>/memory/*.md`	Claude（溢出存储）	主题特定笔记	按需加载
`.claude/rules/*.md`	你（+ `/si:promote` ）	范围规则	打开匹配文件时加载

The promotion lifecycle

推广生命周期

1. Claude discovers pattern → auto-memory (MEMORY.md)
2. Pattern recurs 2-3x → /si:review flags it as promotion candidate
3. You approve → /si:promote graduates it to CLAUDE.md or rules/
4. Pattern becomes an enforced rule, not just a note
5. MEMORY.md entry removed → frees space for new learnings

1. Claude发现模式 → 自动内存（MEMORY.md）
2. 模式重复出现2-3次 → /si:review标记为可推广候选
3. 你确认批准 → /si:promote将其升级至CLAUDE.md或rules/
4. 模式成为强制执行的规则，而非仅笔记
5. MEMORY.md中对应条目移除 → 为新学习内容腾出空间

Core Concepts

核心概念

Auto-memory is capture, not curation

自动内存仅负责捕获，不负责整理

Auto-memory is excellent at recording what Claude learns. But it has no judgment about:

Which learnings are temporary vs. permanent
Which patterns should become enforced rules
When the 200-line limit is wasting space on stale entries
Which solutions are good enough to become reusable skills

That's what this plugin does.

自动内存擅长记录Claude的学习内容，但无法判断：

哪些学习内容是临时的，哪些是永久的
哪些模式应成为强制执行的规则
200行限制下，过时条目是否占用空间
哪些解决方案足够优秀，可成为可复用Skill

而这些正是本插件的功能。

Promotion = graduation

推广 = 升级

When you promote a learning, it moves from Claude's scratchpad (MEMORY.md) to your project's rule system (CLAUDE.md or

.claude/rules/

). The difference matters:

MEMORY.md: "I noticed this project uses pnpm" (background context)
CLAUDE.md: "Use pnpm, not npm" (enforced instruction)

Promoted rules have higher priority and load in full (not truncated at 200 lines).

当您推广某一学习内容时，它会从Claude的草稿本（MEMORY.md）转移至项目的规则系统（CLAUDE.md或

.claude/rules/

）。二者差异显著：

MEMORY.md：“我注意到该项目使用pnpm”（背景上下文）
CLAUDE.md：“使用pnpm，而非npm”（强制执行指令）

已推广的规则优先级更高，且会完整加载（不会被截断为200行）。

Rules directory for scoped knowledge

规则目录用于范围化知识

Not everything belongs in CLAUDE.md. Use

.claude/rules/

for patterns that only apply to specific file types:

yaml

undefined

并非所有内容都适合放在CLAUDE.md中。对于仅适用于特定文件类型的模式，可使用

.claude/rules/

yaml

undefined

.claude/rules/api-testing.md

paths:

"src/api/**/*.test.ts"
"tests/api/**/*"

Use supertest for API endpoint testing
Mock external services with msw
Always test error responses, not just happy paths


This loads only when Claude works with API test files — zero overhead otherwise.

paths:

"src/api/**/*.test.ts"
"tests/api/**/*"

使用supertest进行API端点测试
使用msw模拟外部服务
始终测试错误响应，而非仅测试正常路径


仅当Claude处理API测试文件时，该规则才会加载——其他场景无额外开销。

Agents

代理模块

memory-analyst

Analyzes MEMORY.md and topic files to identify:

Entries that recur across sessions (promotion candidates)
Stale entries referencing deleted files or old patterns
Related entries that should be consolidated
Gaps between what MEMORY.md knows and what CLAUDE.md enforces

分析MEMORY.md和主题文件，识别：

跨会话重复出现的条目（可推广候选）
引用已删除文件或旧模式的过时条目
可合并的相关条目
MEMORY.md记录内容与CLAUDE.md强制执行规则之间的差距

skill-extractor

Takes a proven pattern and generates a complete skill:

SKILL.md with proper frontmatter
Reference documentation
Examples and edge cases
Ready for
```
/plugin install
```
or
```
clawhub publish
```

将已验证的模式生成完整的Skill：

包含正确前置内容的SKILL.md
参考文档
示例和边缘情况
可直接用于
```
/plugin install
```
或
```
clawhub publish
```

Hooks

钩子函数

error-capture (PostToolUse → Bash)

Monitors command output for errors. When detected, appends a structured entry to auto-memory with:

The command that failed
Error output (truncated)
Timestamp and context
Suggested category

Token overhead: Zero on success. ~30 tokens only when an error is detected.

监控命令输出中的错误。检测到错误时，会向自动内存添加结构化条目，包含：

执行失败的命令
错误输出（已截断）
时间戳和上下文
建议分类

Token开销： 执行成功时为0。仅检测到错误时产生约30个Token的开销。

Platform Support

平台支持

Platform	Memory System	Plugin Works?
Claude Code	Auto-memory (MEMORY.md)	✅ Full support
OpenClaw	workspace/MEMORY.md	✅ Adapted (reads workspace memory)
Codex CLI	AGENTS.md	✅ Adapted (reads AGENTS.md patterns)
GitHub Copilot	`.github/copilot-instructions.md`	⚠️ Manual promotion only

平台	内存系统	插件是否可用
Claude Code	自动内存（MEMORY.md）	✅ 完全支持
OpenClaw	workspace/MEMORY.md	✅ 适配支持（读取工作区内存）
Codex CLI	AGENTS.md	✅ 适配支持（读取AGENTS.md中的模式）
GitHub Copilot	`.github/copilot-instructions.md`	⚠️ 仅支持手动推广

self-improving-agent

Original

Translation

Self-Improving Agent

自改进代理（Self-Improving Agent）

Quick Reference

快速参考

How It Fits Together

协作机制

Installation

安装方法

Claude Code (Plugin)

Claude Code（插件）

OpenClaw

OpenClaw

Codex CLI

Codex CLI

Memory Architecture

内存架构

Where things live

文件存储位置

The promotion lifecycle

推广生命周期

Core Concepts

核心概念

Auto-memory is capture, not curation

自动内存仅负责捕获，不负责整理

Promotion = graduation

推广 = 升级

Rules directory for scoped knowledge

规则目录用于范围化知识

.claude/rules/api-testing.md

.claude/rules/api-testing.md

Agents

代理模块

memory-analyst

memory-analyst

skill-extractor

skill-extractor

Hooks

钩子函数

error-capture (PostToolUse → Bash)

error-capture (PostToolUse → Bash)

Platform Support

平台支持

Related

相关资源