baneeishaque/ljt-520_openclaw-backup / skills/0range-x/triple-layer-memory
triple-layer-memory
Verified| File | Kind | Size |
|---|---|---|
| README.md | markdown | 2.1 KB |
| SKILL.md | skill md | 7.3 KB |
| _meta.json | config | 299 B |
| docs/github-upload-guide.md | markdown | 3.1 KB |
| docs/mem0-channel-isolation.md | markdown | 2.2 KB |
| package.json | manifest | 598 B |
| scripts/auto_memory_write.py | script | 4.2 KB |
| scripts/channel_memory.py | script | 5.0 KB |
| scripts/init.sh | script | 4.0 KB |
| scripts/memory_consistency.py | script | 1.9 KB |
| scripts/memory_decay.py | script | 8.9 KB |
| scripts/memory_meta.py | script | 3.0 KB |
| scripts/session_compress.py | script | 3.0 KB |
| templates/MEMORY_ARCHITECTURE.md | markdown | 9.8 KB |
| upload-to-github.sh | script | 2.7 KB |
docs/mem0-channel-isolation.md
View on GitHub# Mem0 频道级命名空间隔离配置
## 概述
为了实现频道级记忆隔离,需要修改 Mem0 插件的 userId 生成逻辑:
- boss 频道:使用主 userId,可以访问全量记忆
- 子频道:使用 `userId::channelKey`,只能访问本频道记忆
## 配置步骤
### 1. 编辑 Mem0 插件
文件位置:`~/.openclaw/extensions/openclaw-mem0/index.ts`
找到 userId 生成逻辑,修改为:
```typescript
// 原代码
const userId = config.userId || "default";
// 修改为
const channelKey = context.channelKey || "boss";
const userId = channelKey === "boss"
? (config.userId || "default")
: `${config.userId || "default"}::${channelKey}`;
```
### 2. 重启 OpenClaw Gateway
```bash
openclaw gateway restart
```
### 3. 验证
在不同频道测试记忆写入和检索:
```bash
# 在 boss 频道
memory_store("这是 boss 频道的记忆")
# 在子频道(如 ca)
memory_store("这是 ca 频道的记忆")
# 在 boss 频道搜索
memory_search("记忆") # 应该能搜到两条
# 在 ca 频道搜索
memory_search("记忆") # 只能搜到 ca 频道的记忆
```
## 文件层隔离
文件层的频道隔离通过元数据 `channel` 字段实现:
```markdown
<!-- meta: importance=7 access=0 created=2026-03-04 last_accessed=2026-03-04 channel=ca -->
```
检索时使用 `scripts/channel_memory.py` 过滤:
```python
from channel_memory import filter_by_channel
entries = read_all_memories()
ca_entries = filter_by_channel(entries, "ca")
```
## 权限边界
- **boss 频道**:全量记忆检索权限
- **子频道**:只能读取本频道记忆
- **跨频道操作**:必须通过 `scripts/channel_guard.py` 鉴权
## 故障排查
### 子频道能看到其他频道的记忆
检查:
1. Mem0 插件的 userId 是否正确生成(应该是 `userId::channelKey`)
2. 文件层的 `channel` 元数据是否正确
3. 检索时是否使用了 `filter_by_channel`
### boss 频道看不到子频道的记忆
检查:
1. boss 频道的 userId 是否是主 userId(不带 `::channelKey`)
2. 检索逻辑是否正确(boss 应该能访问所有频道)
### 记忆写入时没有 channel 字段
检查:
1. `scripts/auto_memory_write.py` 是否传入了 `channel` 参数
2. 元数据模板是否包含 `channel` 字段
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |