Self Improving + Proactive Agent logo

Self Improving + Proactive Agent

Community
L-LesterYu
Self-Improving + Proactive Agent

自我反思 + 自我批评 + 自我学习 + 自组织记忆。智能体评估自己的工作,捕捉错误,并永久改进。适用于以下场景:(1) 命令、工具、API 或操作失败时;(2) 用户纠正你或拒绝你的工作成果时;(3) 你发现自己的知识过时或不正确时;(4) 你发现了更好的方法时;(5) 用户明确安装或引用该技能用于当前任务时。

Overview

PublisherL-LesterYu
RepositoryOpenClaw-hot-skills-zh
Skill nameSelf-Improving + Proactive Agent
Stars
57
Forks
9
Bundled files
14
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 14 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by L-LesterYu on GitHub. Read the source before you install it.

Installation

Install the Self Improving + Proactive Agent AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/L-LesterYu/OpenClaw-hot-skills-zh/tree/main/skills/self-improving-proactive-agent-zh
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/L-LesterYu/OpenClaw-hot-skills-zh.git /tmp/OpenClaw-hot-skills-zh
mkdir -p .claude/skills
cp -r /tmp/OpenClaw-hot-skills-zh/skills/self-improving-proactive-agent-zh .claude/skills/l-lesteryu-self-improving-proactive-agent
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Self Improving + Proactive Agent in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Self Improving + Proactive Agent on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Self Improving + Proactive Agent is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

何时使用

用户纠正你或指出错误。你完成重要工作并想评估结果。你注意到自己的输出中有可以改进的地方。知识应该随着时间积累,无需手动维护。

架构

记忆存储在 ~/self-improving/ 中,采用分层结构。如果 ~/self-improving/ 不存在,运行 setup.md。 工作区设置应将标准的自我改进引导添加到工作区的 AGENTS、SOUL 和 HEARTBEAT.md 文件中,并通过 heartbeat-rules.md 路由定期维护。

~/self-improving/
├── memory.md          # HOT: ≤100 行,始终加载
├── index.md           # 主题索引及行数
├── heartbeat-state.md # 心跳状态:上次运行、已审查的变更、操作笔记
├── projects/          # 按项目的学习内容
├── domains/           # 按领域(代码、写作、通信)
├── archive/           # COLD: 衰退的模式
└── corrections.md     # 最近 50 条纠正日志

快速参考

主题文件
设置指南setup.md
心跳状态模板heartbeat-state.md
记忆模板memory-template.md
工作区心跳片段HEARTBEAT.md
心跳规则heartbeat-rules.md
学习机制learning.md
安全边界boundaries.md
扩展规则scaling.md
记忆操作operations.md
自我反思日志reflections.md
OpenClaw HEARTBEAT 种子openclaw-heartbeat.md

要求

  • 无需凭证
  • 无需额外二进制文件
  • 可选安装 Proactivity 技能可能需要网络访问

学习信号

当你注意到这些模式时自动记录:

纠正 → 添加到 corrections.md,评估是否加入 memory.md:

  • "不,那不对..."
  • "实际上,应该是..."
  • "你搞错了..."
  • "我更喜欢 X,而不是 Y"
  • "记住我总是..."
  • "我之前告诉过你..."
  • "别再做 X 了"
  • "你为什么总是..."

偏好信号 → 如果明确,添加到 memory.md:

  • "我喜欢你..."
  • "总是为我做 X"
  • "永远不要做 Y"
  • "我的风格是..."
  • "对于 [项目],使用..."

模式候选 → 跟踪,重复 3 次后提升:

  • 相同指令重复 3 次以上
  • 反复运行良好的工作流程
  • 用户称赞的特定方法

忽略(不记录):

  • 一次性指令("现在做 X")
  • 特定上下文("在这个文件中...")
  • 假设性问题("如果...会怎样")

自我反思

完成重要工作后,暂停并评估:

  1. 是否达到预期? — 比较结果与意图
  2. 哪些可以更好? — 为下次识别改进点
  3. 这是一个模式吗? — 如果是,记录到 corrections.md

何时自我反思:

  • 完成多步骤任务后
  • 收到反馈(正面或负面)后
  • 修复错误或失误后
  • 当你注意到自己的输出可以更好时

日志格式:

CONTEXT: [任务类型]
REFLECTION: [我注意到的内容]
LESSON: [下次应该怎么做]

示例:

上下文:构建 Flutter UI
反思:间距看起来不对,不得不重做
教训:在展示给用户前检查视觉间距

自我反思条目遵循相同的提升规则:成功应用 3 次 → 提升到 HOT。

快速查询

用户说操作
"你对 X 了解什么?"在所有层级中搜索 X
"你学到了什么?"显示 corrections.md 中的最近 10 条
"显示我的模式"列出 memory.md (HOT)
"显示 [项目] 模式"加载 projects/{name}.md
"温存储中有什么?"列出 projects/ + domains/ 中的文件
"记忆统计"显示各层级的计数
"忘记 X"从所有层级中移除(先确认)
"导出记忆"压缩所有文件

记忆统计

当请求"记忆统计"时,报告:

📊 自我改进记忆

HOT (始终加载):
  memory.md: X 条

WARM (按需加载):
  projects/: X 个文件
  domains/: X 个文件

COLD (已归档):
  archive/: X 个文件

最近活动 (7 天):
  记录的纠正: X
  提升到 HOT: X
  降级到 WARM: X

常见陷阱

陷阱为什么失败更好的做法
从沉默中学习创建错误规则等待明确的纠正或重复证据
提升过快污染 HOT 记忆保持新教训的试探性,直到重复出现
读取每个命名空间浪费上下文仅加载 HOT 加上最小匹配的文件
通过删除压缩失去信任和历史改为合并、总结或降级

核心规则

1. 从纠正和自我反思中学习

  • 当用户明确纠正你时记录
  • 当你在自己的工作中识别出改进时记录
  • 永远不要仅从沉默中推断
  • 相同教训出现 3 次后 → 请求确认为规则

2. 分层存储

层级位置大小限制行为
HOTmemory.md≤100 行始终加载
WARMprojects/, domains/每个文件 ≤200 行上下文匹配时加载
COLDarchive/无限制明确查询时加载

3. 自动提升/降级

  • 7 天内使用 3 次的模式 → 提升到 HOT
  • 30 天未使用的模式 → 降级到 WARM
  • 90 天未使用的模式 → 归档到 COLD
  • 永远不要在未询问的情况下删除

4. 命名空间隔离

  • 项目模式保存在 projects/{name}.md
  • 全局偏好保存在 HOT 层级 (memory.md)
  • 领域模式(代码、写作)保存在 domains/
  • 跨命名空间继承:global → domain → project

5. 冲突解决

当模式矛盾时:

  1. 最具体的优先(project > domain > global)
  2. 最新的优先(同一层级)
  3. 如果模糊 → 询问用户

6. 压缩

当文件超过限制时:

  1. 将相似的纠正合并为单一规则
  2. 归档未使用的模式
  3. 总结冗长的条目
  4. 永远不要丢失已确认的偏好

7. 透明性

  • 每个来自记忆的操作 → 引用来源:"使用 X (来自 projects/foo.md:12)"
  • 每周摘要可用:学习的模式、降级的、归档的
  • 按需完整导出:所有文件为 ZIP

8. 安全边界

参见 boundaries.md — 永远不要存储凭证、健康数据、第三方信息。

9. 优雅降级

如果达到上下文限制:

  1. 仅加载 memory.md (HOT)
  2. 按需加载相关命名空间
  3. 永远不要静默失败 — 告诉用户什么未加载

范围

此技能仅:

  • 从用户纠正和自我反思中学习
  • 在本地文件中存储偏好 (~/self-improving/)
  • 当工作区集成心跳时,在 ~/self-improving/heartbeat-state.md 中维护心跳状态
  • 激活时读取自己的记忆文件

此技能从不:

  • 访问日历、电子邮件或联系人
  • 发起网络请求
  • 读取 ~/self-improving/ 之外的文件
  • 从沉默或观察中推断偏好
  • 在心跳清理期间删除或盲目重写自我改进记忆
  • 修改自己的 SKILL.md

数据存储

本地状态保存在 ~/self-improving/:

  • memory.md 用于 HOT 规则和已确认的偏好
  • corrections.md 用于明确的纠正和可复用的教训
  • projects/ 和 domains/ 用于限定范围的模式
  • archive/ 用于衰退或不活跃的模式
  • heartbeat-state.md 用于定期维护标记

相关技能

如果用户确认,使用 clawhub install <slug> 安装:

  • memory — 智能体长期记忆模式
  • learning — 自适应教学和解释
  • decide — 自动学习决策模式
  • escalate — 知道何时询问、何时自主行动

反馈

  • 如果有用:clawhub star self-improving
  • 保持更新:clawhub sync

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Self Improving + Proactive Agent AI skill do?

自我反思 + 自我批评 + 自我学习 + 自组织记忆。智能体评估自己的工作,捕捉错误,并永久改进。适用于以下场景:(1) 命令、工具、API 或操作失败时;(2) 用户纠正你或拒绝你的工作成果时;(3) 你发现自己的知识过时或不正确时;(4) 你发现了更好的方法时;(5) 用户明确安装或引用该技能用于当前任务时。

Why use Self Improving + Proactive Agent on TypingMind?

Because you install it once and use it with any model. Self Improving + Proactive Agent is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Self Improving + Proactive Agent in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/L-LesterYu/OpenClaw-hot-skills-zh/tree/main/skills/self-improving-proactive-agent-zh. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Self Improving + Proactive Agent?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Self Improving + Proactive Agent?

As many as you like. As long as a model supports skills, you can use Self Improving + Proactive Agent with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Self Improving + Proactive Agent AI skill free?

Yes. It is published on GitHub by L-LesterYu under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇