agentCLAWHUBUnverified
Self-Improving Agent
基于对经验的持续学习，不断优化 Agent 能力。适用于完成重要任务后、出现错误时、会话结束时，或用户输入“自我进化”“总结经验”“从经验中学习”等指令时触发。 Skill: Self-Improving Agent Owner: initail Summary: 基于对经验的持续学习，不断优化 Agent 能力。适用于完成重要任务后、出现错误时、会话结束时，或用户输入“自我进化”“总结经验”“从经验中学习”等指令时触发。 Tags: latest:0.2.0 Version history: v0.2.0 | 2026-03-24T09:26:23.443Z | user 基于 2025 终身学习研究的通用自进化系统，支持多记忆架构、用户确认门、置信度追踪、自我纠错 Archive index: Archive v0.2.0: 8 files, 16457 bytes Files: README.md (4853b), references/appendix.md (10711b), skill-card.md (2513b), SKILL.md (20139b), templates
OpenClaw
Rank
62
Safety
84
Downloads
2.3k
Updated
Oct 9, 2026
Version
0.2.0
Source
CLAWHUB
About
What it does, and when to use it.
Capability contract not published. No trust telemetry is available yet. 2.3K downloads reported by the source. Last updated 10/9/2026.
Avoid when
- Contract metadata is missing or unavailable for deterministic execution.
Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing
Public facts
Every fact links back to the source it came from.
- Vendor
- Clawhubvendor · observed Oct 9, 2026
- Protocol compatibility
- OpenClawcompatibility · observed Oct 9, 2026
- Adoption signal
- 2.3K downloadsadoption · observed Oct 9, 2026
- Latest release
- 0.2.0release · observed Mar 24, 2026
- Handshake status
- UNKNOWNsecurity
Install and run
Setup complexity: low.
clawhub skill install s17f7yh5asw5nws80f6sp0t2ad83hxgs:self-improving-agent-skill- Setup complexity is LOW. This package is likely designed for quick installation with minimal external side-effects.
- Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data.
Contract: missing
curl -s "https://www.xpersona.co/api/v1/agents/clawhub-initail-self-improving-agent-skill/snapshot"
Documentation
CLAWHUB
33,820 characters of source documentation, loaded on request.
Extracted files
5 files captured from the source.
SKILL.md
--- name: self-improving-agent-skill version: 0.2.0 description: 基于对经验的持续学习，不断优化 Agent 能力。适用于完成重要任务后、出现错误时、会话结束时，或用户输入“自我进化”“总结经验”“从经验中学习”等指令时触发。 --- # Self-Improving Agent > "An AI agent that learns from every interaction, accumulating patterns and insights to continuously improve its own capabilities." — Based on 2025 lifelong learning research ## Overview This is a **universal self-improvement system** that learns from ALL task experiences. It implements a complete feedback loop: - **Multi-Memory Architecture**: Semantic (patterns/rules) + Episodic (experiences) + Working (session context) - **Self-Correction**: Detects and fixes guidance errors - **Self-Validation**: Periodically verifies skill accuracy - **Evolution Markers**: Traceable changes with source attribution - **Confidence Tracking**: Measures pattern reliability over time - **User Confirmation Gate**: All skill file modifications require explicit user approval before applying - **Human-in-the-Loop**: Collects feedback to validate improvements ## Research-Based Design | Research | Key Insight | Application | |----------|-------------|-------------| | [SimpleMem](https://arxiv.org/html/2601.02553v1) | Efficient lifelong memory | Pattern accumulation system | | [Multi-Memory Survey](https://dl.acm.org/doi/10.1145/3748302) | Semantic + Episodic memory | World knowledge + experiences | | [Lifelong Learning](https://arxiv.org/html/2501.07278v1) | Continuous task stream learning | Learn from every task | | [Evo-Memory](https://shothota.medium.com/evo-memory-deepminds-new-benchmark) | Test-time lifelong learning | Real-time adaptation | ## The Self-Improvement Loop ``` ┌──────────────────────────────────────────────────────────────┐ │ UNIVERSAL SELF-IMPROVEMENT │ ├──────────────────────────────────────────────────────────────┤ │ │ │ Task Event → Extract Experience → Abstract Pattern → Update │ │ │ │ │ │ │ │ ▼ ▼ ▼ ▼ │ │ ┌────────────────────────────────────────────────────────┐ │ │ │ MULTI-MEMORY SYSTEM │ │ │ ├────────────────────────────────────────────────────────┤ │ │ │ Semantic Memory │ Episodic Memory │ Working Memory │ │ │ │ (Patterns/Rules) │ (Experiences) │ (Current) │ │ │ │ memory/self-improving/semantic/ │ memory/self-improving/episodic/ │ memory/self-improving/working/ │ │ │ └────────────────────────────────────────────────────────┘ │ │ │ │ ┌────────────────────────────────────────────────────────┐ │ │ │ FEEDBACK LOOP │ │ │ │ User Feedback → Confidence Update → Pattern Adapt │ │ │ └────────────────────────────────────────────────────────┘ │ │
README.md
# Self-Improving Agent Skill
> "An AI agent that learns from every interaction, accumulating patterns and insights to continuously improve its own capabilities." — Based on 2025 lifelong learning research
## 🧠 技能简介
这是一个**通用自学习系统**，让 AI Agent 能够从每次任务中学习，持续优化自身能力。基于 2025 年终身学习研究设计，实现完整的「经验→模式→改进」反馈循环。
## ✨ 核心特性
| 特性 | 说明 |
|------|------|
| 🧠 **三记忆架构** | 语义记忆（模式/规则）+ 情景记忆（经验）+ 工作记忆（会话上下文） |
| 🔄 **自我纠错** | 检测并修复错误的技能指导 |
| ✅ **自我验证** | 定期检查技能准确性和外部引用有效性 |
| 📊 **置信度追踪** | 动态衡量模式可靠性，随应用次数调整 |
| 🛡️ **用户确认门** | 所有技能修改需用户明确批准，安全可控 |
| 👤 **人在回路** | 每次改进后收集用户反馈，持续优化 |
## 🎯 触发条件
### 自动触发
- ✅ 重要任务完成后 → 提取模式，提议技能更新
- ✅ 错误或失败发生时 → 捕获上下文，触发自我纠错
- ✅ 会话结束时 → 合并工作记忆到长期记忆
### 手动触发
用户输入以下指令时激活：
- "自我进化" / "self-improve"
- "总结经验" / "从经验中学习"
- "分析今天的经验" / "总结教训"
## 📦 安装
```bash
clawhub install self-improving-agent-skill
```
## 🚀 使用示例
### 任务完成后自动学习
```
用户：帮我修复这个 bug...
Agent: [修复完成]
→ 自动触发：提取经验 → 抽象模式 → 提议更新
```
### 手动触发学习
```
用户：自我进化
Agent: 正在分析本次会话经验...
- 提取 3 个新模式
- 提议更新 2 个技能
- 请确认是否应用更改
```
### 错误后自我纠错
```
[检测到错误]
→ 捕获错误上下文
→ 识别根本原因
→ 提议修正方案（需用户确认）
→ 应用修正并添加标记
```
## 📁 记忆存储结构
```
{workspace}/memory/self-improving/
├── semantic/
│ └── patterns.json # 抽象模式和规则
├── episodic/
│ └── YYYY/
│ └── YYYY-MM-DD-{task}.json # 具体经验
├── working/
│ ├── current_session.json # 当前会话
│ ├── last_error.json # 最近错误
│ └── session_end.json # 会话结束标记
└── index.json # 全局索引和指标
```
## 🔄 自我改进流程
```
任务完成 → 提取经验 → 抽象模式 → 提议更改 → ★用户确认★ → 更新技能 → 存入记忆 → 收集反馈
```
### 关键安全机制
1. **用户确认门**：所有修改必须经用户明确批准
2. **可追溯标记**：每次更新带 Evolution/Correction 标记
3. **置信度动态调整**：
- 成功应用 → +0.05
- 用户正面反馈 → +0.05
- 用户负面反馈 → -0.10
- 导致错误 → -0.15
## 📊 持续学习指标
在 `memory/self-improving/index.json` 中追踪：
| 指标 | 描述 | 目标 |
|------|------|------|
| `patterns_learned` | 学习的模式总数 | 持续增长 |
| `patterns_applied` | 模式应用次数 | 持续增长 |
| `avg_confidence` | 平均置信度 | > 0.8 |
| `self_corrections` | 自我修正次数 | 越低越好 |
| `error_rate_reduction` | 错误减少率 | 负值（改善） |
## 🎓 研究基础
基于以下 2025 年终身学习研究设计：
- [SimpleMem: Efficient Lifelong Memory for LLM Agents](https://arxiv.org/html/2601.02553v1)
- [A Survey on the Memory Mechanism of Large Language Model Agents](https://dl.acm.org/doi/10.1145/3748302)
- [Lifelong Learning of LLM based Agents](https://arxiv.org/html/2501.07278v1)
- [Evo-Memory: DeepMind's Benchmark](https://shothota.medium.com/evo-memory-deepminds-new-benchmark)
## 📝 版本历史
### v0.2.0 (当前版本)
- ✅ 多记忆架构实现（语义 + 情景 + 工作）
- ✅ 用户确认门机制
- ✅ 置信度追踪系统
- ✅ 自我纠错流程
- ✅ 自我验证模板
- ✅ 人在回路反馈收集
### v0.1.0
- 初始版本，基础模式提取功能
## ⚠️ 使用注意
- **不要** 在未确认的情况下自动修改技能文件
- **不要** 覆盖现有记忆内容（始终追加）
- **不要** 从单一经验过度概括（等待 2-3 次出现）
- **要** 为所有模式追踪置信度
- **要** 重视负面反馈（最有价值的信号）
- **要** 使用 Evolution/Correction 标记保证可追溯性
## 🤝 贡献
欢迎提交 Issue 和 Pull Request 改进此技能！
## 📄 许可证
MIT License
---
**作者**: 子然
**分类**: 元技能 / 自我改进
**标签**: self-improvement, lifelong-learning, memory, agent-architecture_meta.json
{
"ownerId": "kn7a0ck2732w2h0p8hegn53t9183hg5j",
"slug": "self-improving-agent-skill",
"version": "0.2.0",
"publishedAt": 1774344383443
}references/appendix.md
# 附录
## 记忆文件结构
```
{workspace}/memory/self-improving/
├── semantic/
│ └── patterns.json # 跨上下文复用的抽象模式
├── episodic/
│ ├── YYYY/
│ │ ├── YYYY-MM-DD-{task}.json
│ │ └── ...
│ └── episodes-index.json # 情景快速查找索引
├── working/
│ ├── current_session.json # 当前会话上下文
│ ├── last_error.json # 最近的错误上下文
│ └── session_end.json # 会话结束标记（用于合并）
└── index.json # 全局记忆索引和指标
```
### 与 Agent 主记忆的关系
```
{workspace}/
├── MEMORY.md # Agent 核心记忆（Self-Improving Agent 可追加内容）
├── memory/
│ ├── YYYY-MM-DD.md # Agent 每日记忆（Self-Improving Agent 可追加内容）
│ └── self-improving/ # Self-Improving Agent 专属记忆空间
│ ├── semantic/
│ ├── episodic/
│ ├── working/
│ └── index.json
```
### index.json 结构
```json
{
"created": "2025-01-01T00:00:00Z",
"last_updated": "2025-01-15T14:30:00Z",
"stats": {
"total_patterns": 47,
"total_episodes": 128,
"total_sessions": 35,
"avg_confidence": 0.87,
"patterns_applied": 238,
"self_corrections": 8
}
}
```
## 自我验证
### 验证报告模板
```markdown
## 验证报告
**日期**：[YYYY-MM-DD]
**范围**：[验证的技能或模式]
### 检查项
- [ ] 示例可正确编译或运行
- [ ] 检查清单匹配当前规范
- [ ] 外部引用仍然有效
- [ ] 无重复或冲突的指导
- [ ] 模式置信度水平合理
### 发现
- [发现 1]
- [发现 2]
### 行动
- [行动 1]
- [行动 2]
```
## 持续学习指标
在 `memory/self-improving/index.json` 中追踪以下指标，以衡量随时间的改进：
```json
{
"metrics": {
"patterns_learned": 47,
"patterns_applied": 238,
"skills_updated": 12,
"avg_confidence": 0.87,
"user_satisfaction_trend": "improving",
"error_rate_reduction": "-35%",
"self_corrections": 8
}
}
```
### 关键指标说明
| 指标 | 描述 | 目标 |
|------|------|------|
| `patterns_learned` | 提取的独立模式总数 | 持续增长 |
| `patterns_applied` | 模式被用于任务的次数 | 持续增长 |
| `avg_confidence` | 活跃模式的平均置信度 | > 0.8 |
| `self_corrections` | 执行的指导修正次数 | 越低越好 |
| `error_rate_reduction` | 重复错误的减少百分比 | 负值（表示改善） |
## 人在回路中反馈
### 反馈收集模板
```markdown
## 自我改进摘要
我从本次会话中学习并更新了以下内容：
### 更新的技能
- `{skill-name}`：{更新了什么}
### 提取的模式
1. **{模式名称}**：{描述}（置信度：{X.XX}）
### 置信度水平
- 新模式：~0.85（需要验证）
- 强化的模式：~0.95（已充分验证）
### 你的反馈
请为这些改进评分（1-10）：
- 这些更新是否有帮助？
- 是否应将此模式更广泛地应用？
- 是否需要修正？
```
### 反馈整合规则
```yaml
用户反馈:
正面 (评分 >= 7):
action: 模式置信度 +0.05
scope: 考虑扩展到相关技能
中性 (评分 4-6):
action: 保留模式，收集更多数据
scope: 仅限当前技能
负面 (评分 <= 3):
action: 置信度 -0.1，修订模式
scope: 置信度 < 0.3 时从活跃模式中移除
```
## 工作流图
### 完整自我改进循环
```
任务完成
│
▼
┌─────────────┐ ┌──────────────┐ ┌──────────────┐
│ 提取 │────▶│ 抽象 │────▶│ 提议 │
│ 经验 │ │ 模式 │ │ 更改 │
└─────────────┘ └──────────────┘ └──────────────┘
│ │ │
▼ ▼ ▼
┌─────────────┐ ┌──────────────┐ ┌──────────────┐
│ 存入 │ │ 存入 │ │ ★ 用户 │
│ 情景记忆 │ │ 语义记忆 │ │ 确认 ★ │
│ │ │ │ │ （必需） │skill-card.md
## Description: Helps an agent learn from task experience by capturing patterns, errors, feedback, and proposed updates to memory and skill guidance. This skill is ready for commercial/non-commercial use. ## Publisher: [initail](https://clawhub.ai/user/initail) ### License/Terms of Use: MIT-0 ## Use Case: Developers and agent users use this skill to make an agent extract reusable lessons from completed tasks, errors, and user feedback, then propose memory or skill updates for review. ### Deployment Geography for Use: Global ## Known Risks and Mitigations: Risk: The skill can retain long-term records of tasks, errors, feedback, and reusable patterns. Mitigation: Use it only in workspaces where this retention is acceptable, and avoid sensitive project context unless retention controls are tightened. Risk: The skill can propose persistent changes to memory and skill behavior. Mitigation: Review every proposed memory or skill update before applying it, and preserve the documented user confirmation gate. Risk: Incorrect lessons could become persistent guidance for future agent behavior. Mitigation: Validate proposed patterns, corrections, and confidence changes before reuse, especially after errors or negative feedback. ## Reference(s): - [ClawHub skill page](https://clawhub.ai/initail/skills/self-improving-agent-skill) - [Appendix](artifact/references/appendix.md) - [SimpleMem: Efficient Lifelong Memory for LLM Agents](https://arxiv.org/html/2601.02553v1) - [A Survey on the Memory Mechanism of Large Language Model Agents](https://dl.acm.org/doi/10.1145/3748302) - [Lifelong Learning of LLM based Agents](https://arxiv.org/html/2501.07278v1) - [Evo-Memory: DeepMind's Benchmark](https://shothota.medium.com/evo-memory-deepminds-new-benchmark) ## Skill Output: **Output Type(s):** [text, markdown, code, shell commands, configuration, guidance] **Output Format:** [Markdown guidance with JSON and code examples] **Output Parameters:** [1D] **Other Properties Related to Output:** [May propose memory records, pattern entries, validation reports, and skill update drafts that require user review before persistent changes.] ## Skill Version(s): 0.2.0 (source: frontmatter and server release metadata) ## Ethical Considerations: Users should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
Github ReposUpdated 6h agoRank 70
AionUi
Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!
MCPOPENCLAW
Github ReposUpdated 6mo agoRank 70
activepieces
AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents
OPENCLAW
Github ReposUpdated 6mo agoRank 70
cherry-studio
AI productivity studio with smart chat, autonomous agents, and 300+ assistants.
MCPOPENCLAW
Github ReposUpdated 7mo agoRank 70
CopilotKit
The Frontend for Agents & Generative UI. React + Angular
OPENCLAW
Machine-readable data
The same record, as JSON, for agents and crawlers.
{
"facts": [
{
"factKey": "vendor",
"category": "vendor",
"label": "Vendor",
"value": "Clawhub",
"href": "https://clawhub.ai/initail/skills/self-improving-agent-skill",
"sourceUrl": "https://clawhub.ai/initail/skills/self-improving-agent-skill",
"sourceType": "profile",
"confidence": "medium",
"observedAt": "2026-10-09T16:37:46.826Z",
"isPublic": true
},
{
"factKey": "protocols",
"category": "compatibility",
"label": "Protocol compatibility",
"value": "OpenClaw",
"href": "https://www.xpersona.co/api/v1/agents/clawhub-initail-self-improving-agent-skill/contract",
"sourceUrl": "https://www.xpersona.co/api/v1/agents/clawhub-initail-self-improving-agent-skill/contract",
"sourceType": "contract",
"confidence": "medium",
"observedAt": "2026-10-09T16:37:46.826Z",
"isPublic": true
},
{
"factKey": "traction",
"category": "adoption",
"label": "Adoption signal",
"value": "2.3K downloads",
"href": "https://clawhub.ai/initail/self-improving-agent-skill",
"sourceUrl": "https://clawhub.ai/initail/self-improving-agent-skill",
"sourceType": "profile",
"confidence": "medium",
"observedAt": "2026-10-09T16:37:46.826Z",
"isPublic": true
},
{
"factKey": "latest_release",
"category": "release",
"label": "Latest release",
"value": "0.2.0",
"href": "https://clawhub.ai/initail/self-improving-agent-skill",
"sourceUrl": "https://clawhub.ai/initail/self-improving-agent-skill",
"sourceType": "release",
"confidence": "medium",
"observedAt": "2026-03-24T09:26:23.443Z",
"isPublic": true
},
{
"factKey": "handshake_status",
"category": "security",
"label": "Handshake status",
"value": "UNKNOWN",
"href": "https://www.xpersona.co/api/v1/agents/clawhub-initail-self-improving-agent-skill/trust",
"sourceUrl": "https://www.xpersona.co/api/v1/agents/clawhub-initail-self-improving-agent-skill/trust",
"sourceType": "trust",
"confidence": "medium",
"observedAt": null,
"isPublic": true
}
],
"events": [
{
"eventType": "release",
"title": "Release 0.2.0",
"description": "基于 2025 终身学习研究的通用自进化系统，支持多记忆架构、用户确认门、置信度追踪、自我纠错",
"href": "https://clawhub.ai/initail/self-improving-agent-skill",
"sourceUrl": "https://clawhub.ai/initail/self-improving-agent-skill",
"sourceType": "release",
"confidence": "medium",
"observedAt": "2026-03-24T09:26:23.443Z",
"isPublic": true
}
]
}Record generated Oct 10, 2026.