According to AlphaMoat, as of 2026-09, 总结 has 3,186 downloads and 3 stars — ranked #3,580 of 5,078 AI skills overall, and #307 of 442 in Knowledge.

Installs

155

Downloads

3.2K

Stars

Overall Rank

#3,580

Top 71% overall · Top 70% in Knowledge

Platform	Installs	Downloads	Stars	Comments
ClawHub ↗	45	1.5K	-	-
SkillHub ↗	155	3.2K	3	—

Platforms count differently and update at different times.

What this skill does

It helps decide whether fixes, configuration changes or key decisions should be saved, then organize root cause, steps and lessons. Simple questions and casual chat are excluded. Persistent memory may include sensitive details or instruction-like text; review and sanitize it, and confirm priority, category and sharing scope before saving. Check existing records for duplicates before merging a lesson, and confirm the exact memory layer and file. Optional index/query commands require the separately available memory_mgr.py and Python; the document alone does not prove retention cleanup or indexing is configured. P0/L0 or shared records can affect later sessions, so explicitly review that scope. Commands inside incident examples are historical content, not actions to rerun automatically.

Profile published 2026-10-27

What you can do with it

Record troubleshooting or configuration-change lessons
Organize reusable operational notes

When to use it

When information is worth retaining and secrets, private links and untrusted instructions have been removed

Example prompts

Generated from the skill’s documentation as examples; not results of actual runs.

Draft a non-sensitive incident review without writing it to memory.

List checks before saving content to persistent memory.

Monthly Trend

Downloads3.2K

Stars3

The counting method changed in 2026-09: skills with the same name are now counted separately by author, and the full catalog is covered. Dashed lines show months before the change; they are not compared with later months, and no growth is computed across the change.

Month	Downloads	MoM	Stars
2026-09	3.2K	—	3
2026-08	2.3K	+36.5%	3
2026-07	1.7K	+36.3%	2
2026-06	1.2K	+18.0%	2
2026-05	1.1K	—	2
2026-03	638	—	-

SKILL.md (original)

The author’s original file, shown as plain text without edits or review by AlphaMoat. Links are not clickable; check commands before running them. Fetched 2026-10-23.

3.7 KB · ClawHub v1.0.5

# Skill: 总结 (Zongjie)

将重要事件整理成结构化文档，沉淀经验到记忆系统。

## 记忆管理系统

基于 OpenViking 方案的三层记忆系统：

| 层级 | 说明 | 加载时机 |
|------|------|----------|

Show all (172 lines)

| **L0** | 核心索引，始终加载 | 每次会话 |
| **L1** | 重要记忆，按需加载 | 会话开始时 |
| **L2** | 归档记忆，查询加载 | 显式查询时 |

| 优先级 | 生命周期 | 说明 |
|--------|----------|------|
| **P0** | 永不过期 | 核心人设、操作规则、关键偏好 |
| **P1** | 90天 | 重要经验、解决方案 |
| **P2** | 30天 | 临时记录，自动清理 |

## 使用场景

- 系统故障排查完成
- 有参考价值的调试过程
- 需要记录解决方案供以后查阅
- **不保存**：简单的问答、查询、闲聊

## 精简模板

```markdown
# 标题

> P1 | L1

### 问题
- [简明描述]

### 根因
- [根本原因]

### 解决步骤
1. 步骤1
2. 步骤2
3. 步骤3

### 关键信息
- 报错信息/日志
- 关键命令
- 参考链接

### 结果
- ✅/❌ 状态

### 经验
- 关键点
```

## 步骤

### M1: 判断价值

以下情况**不保存**：
- 简单问答（如"这个公司怎么样"）
- 查询类（如"帮我查下天气"）
- 闲聊

以下情况**需要保存**：
- 系统故障及修复
- 配置变更
- 重要决策
- 调试过程

### M2: 确定优先级

根据内容判断 P/L 级别：

| 分类 | P级 | L级 | 示例 |
|------|-----|-----|------|
| 人设/操作规则 | P0 | L0 | 核心规则、沟通原则 |
| 系统故障修复 | P1 | L1 | Gateway崩溃、Cron失败 |
| 新功能配置 | P1 | L1 | GitHub同步、自动备份 |
| 调试过程 | P2 | L2 | 排查步骤、临时方案 |

### M3: 详细整理

提取核心信息：
- 问题是什么
- 根因是什么
- **解决步骤**（一步一步写清楚）
- 关键信息（报错日志、命令、链接）
- 结果如何
- 经验教训（1-2条）

### M4: 合并同类项

检查现有记录：
- 如果已有相同/类似问题 → 合并到已有记录中
- 更新"解决步骤"部分
- 更新"经验"部分

### M5: 保存分类

**分类文件**：
- **系统问题** → `memory/topics/03-系统调试.md`
- **操作规则** → `memory/topics/02-操作规则.md`
- **人设偏好** → `memory/topics/01-人设.md`
- **新功能** → `memory/topics/04-XXX.md`
- **其他** → `memory/YYYY-MM-DD.md`

**索引更新** (可选)：
```bash
# 更新索引
python3 ~/.openclaw/scripts/memory_mgr.py add <agent> "<内容>" P1 L1
```

### M6: 输出确认

告诉用户保存位置和 P/L 分类。

## 示例

用户："帮我记录今天 Cron 推送失败的排查"

→ 检查发现已有类似记录 → 合并更新：
```
# Cron 推送失败修复

> P1 | L1

### 问题
- Cron 早间推送失败 (announce delivery failed)

### 根因
- .openclaw/openclaw.json 有 JSON 尾随逗号

### 解决步骤
1. 检查 cron 状态: `openclaw cron list`
2. 手动测试 Telegram: `openclaw message send ...`
3. 检查 JSON: `python3 -m json.tool ~/.openclaw/openclaw.json`
4. 修复 JSON 语法错误
5. 重新安装: `npm install -g openclaw`
6. 重启: `openclaw gateway restart`

### 关键信息
- 报错: "cron announce delivery failed"
- 日志: `openclaw logs --limit 50`
- 配置: `~/.openclaw/openclaw.json`

### 结果
- ✅ 推送恢复正常

### 经验
- JSON 错误会导致投递失败
- 重新安装解决模块丢失
```

## 常用命令

```bash
# 添加记忆 (通过脚本)
python3 ~/.openclaw/scripts/memory_mgr.py add work "内容" P1 L1

# 查询记忆
python3 ~/.openclaw/scripts/memory_mgr.py query work

# 获取 L0 摘要
python3 ~/.openclaw/scripts/memory_mgr.py layer work L0

# 共享记忆
python3 ~/.openclaw/scripts/memory_mgr.py shared add "项目规范" "..."
```