memory-setup

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Memory Setup Skill

记忆设置技能

Transform your agent from goldfish to elephant. This skill helps configure persistent memory for Moltbot/Clawdbot.

将你的Agent从只有7秒记忆的金鱼变成过目不忘的大象，该技能可帮助你为Moltbot/Clawdbot配置持久化记忆。

Quick Setup

快速设置

1. Enable Memory Search in Config

1. 在配置文件中启用记忆搜索

Add to

~/.clawdbot/clawdbot.json

(or

moltbot.json

json

{
  "memorySearch": {
    "enabled": true,
    "provider": "voyage",
    "sources": ["memory", "sessions"],
    "indexMode": "hot",
    "minScore": 0.3,
    "maxResults": 20
  }
}

添加以下配置到

~/.clawdbot/clawdbot.json

（或

moltbot.json

）中：

json

{
  "memorySearch": {
    "enabled": true,
    "provider": "voyage",
    "sources": ["memory", "sessions"],
    "indexMode": "hot",
    "minScore": 0.3,
    "maxResults": 20
  }
}

2. Create Memory Structure

2. 创建记忆目录结构

In your workspace, create:

workspace/
├── MEMORY.md              # Long-term curated memory
└── memory/
    ├── logs/              # Daily logs (YYYY-MM-DD.md)
    ├── projects/          # Project-specific context
    ├── groups/            # Group chat context
    └── system/            # Preferences, setup notes

在你的工作区中创建如下目录结构：

workspace/
├── MEMORY.md              # 长期维护的记忆文件
└── memory/
    ├── logs/              # 日常日志（命名格式为YYYY-MM-DD.md）
    ├── projects/          # 项目专属上下文
    ├── groups/            # 群组聊天上下文
    └── system/            # 偏好设置、安装记录

3. Initialize MEMORY.md

3. 初始化MEMORY.md

Create

MEMORY.md

in workspace root:

markdown

undefined

在工作区根目录创建

MEMORY.md

文件：

markdown

undefined

MEMORY.md — Long-Term Memory

MEMORY.md — 长期记忆

About [User Name]

关于[用户名]

Key facts, preferences, context

关键信息、偏好、背景上下文

Active Projects

进行中的项目

Project summaries and status

项目概述和进度

Decisions & Lessons

决策与经验教训

Important choices made
Lessons learned

已做出的重要选择
总结的经验教训

Preferences

偏好设置

Communication style
Tools and workflows

undefined

沟通风格
工具和工作流偏好

undefined

Config Options Explained

配置选项说明

Setting	Purpose	Recommended
`enabled`	Turn on memory search	`true`
`provider`	Embedding provider	`"voyage"`
`sources`	What to index	`["memory", "sessions"]`
`indexMode`	When to index	`"hot"` (real-time)
`minScore`	Relevance threshold	`0.3` (lower = more results)
`maxResults`	Max snippets returned	`20`

设置项	用途	推荐值
`enabled`	开启记忆搜索	`true`
`provider`	嵌入模型提供商	`"voyage"`
`sources`	要索引的内容来源	`["memory", "sessions"]`
`indexMode`	索引时机	`"hot"` （实时索引）
`minScore`	相关性阈值	`0.3` （值越低返回结果越多）
`maxResults`	返回的最大片段数	`20`

Provider Options

提供商可选值

```
voyage
```
— Voyage AI embeddings (recommended)
```
openai
```
— OpenAI embeddings
```
local
```
— Local embeddings (no API needed)

```
voyage
```
— Voyage AI嵌入模型（推荐）
```
openai
```
— OpenAI嵌入模型
```
local
```
— 本地嵌入模型（无需API密钥）

Source Options

来源可选值

```
memory
```
— MEMORY.md + memory/*.md files
```
sessions
```
— Past conversation transcripts
```
both
```
— Full context (recommended)

```
memory
```
— MEMORY.md + memory目录下所有.md文件
```
sessions
```
— 过往对话记录
```
both
```
— 全部上下文（推荐）

Daily Log Format

日常日志格式

Create

memory/logs/YYYY-MM-DD.md

daily:

markdown

undefined

每日创建

memory/logs/YYYY-MM-DD.md

文件，格式如下：

markdown

undefined

YYYY-MM-DD — Daily Log

YYYY-MM-DD — 日常日志

[Time] — [Event/Task]

[时间] — [事件/任务]

What happened
Decisions made
Follow-ups needed

发生的内容
做出的决策
需要跟进的事项

[Time] — [Another Event]

[时间] — [其他事件]

Details

undefined

详情

undefined

Agent Instructions (AGENTS.md)

Agent使用说明（AGENTS.md）

Add to your AGENTS.md for agent behavior:

markdown

undefined

添加以下内容到AGENTS.md中定义Agent行为：

markdown

undefined

Memory Recall

记忆召回

Before answering questions about prior work, decisions, dates, people, preferences, or todos:

Run memory_search with relevant query
Use memory_get to pull specific lines if needed
If low confidence after search, say you checked

undefined

在回答关于过往工作、决策、日期、人物、偏好或待办的问题前：

使用相关关键词运行memory_search
必要时使用memory_get拉取特定行内容
如果搜索后置信度较低，告知用户你已经查询过记忆库

undefined

Troubleshooting

故障排查

Memory search not working?

记忆搜索不工作？

Check
```
memorySearch.enabled: true
```
in config
Verify MEMORY.md exists in workspace root
Restart gateway:
```
clawdbot gateway restart
```

检查配置中
```
memorySearch.enabled: true
```
是否设置正确
确认工作区根目录存在MEMORY.md文件
重启网关：
```
clawdbot gateway restart
```

Results not relevant?

搜索结果不相关？

Lower
```
minScore
```
to
```
0.2
```
for more results
Increase
```
maxResults
```
to
```
30
```
Check that memory files have meaningful content

将
```
minScore
```
调低到
```
0.2
```
获取更多结果
将
```
maxResults
```
调高到
```
30
```
检查记忆文件内容是否有实际意义

Provider errors?

提供商报错？

Voyage: Set
```
VOYAGE_API_KEY
```
in environment
OpenAI: Set
```
OPENAI_API_KEY
```
in environment
Use
```
local
```
provider if no API keys available

Voyage：在环境变量中设置
```
VOYAGE_API_KEY
```
OpenAI：在环境变量中设置
```
OPENAI_API_KEY
```
如果没有API密钥，可以使用
```
local
```
提供商

Verification

验证

Test memory is working:

User: "What do you remember about [past topic]?"
Agent: [Should search memory and return relevant context]

If agent has no memory, config isn't applied. Restart gateway.

测试记忆功能是否正常工作：

用户：「你记得[过往话题]的相关内容吗？」
Agent：[应该搜索记忆库并返回相关上下文]

如果Agent没有相关记忆，说明配置未生效，请重启网关。

Full Config Example

完整配置示例

json

{
  "memorySearch": {
    "enabled": true,
    "provider": "voyage",
    "sources": ["memory", "sessions"],
    "indexMode": "hot",
    "minScore": 0.3,
    "maxResults": 20
  },
  "workspace": "/path/to/your/workspace"
}

json

{
  "memorySearch": {
    "enabled": true,
    "provider": "voyage",
    "sources": ["memory", "sessions"],
    "indexMode": "hot",
    "minScore": 0.3,
    "maxResults": 20
  },
  "workspace": "/path/to/your/workspace"
}

Why This Matters

使用意义

Without memory:

Agent forgets everything between sessions
Repeats questions, loses context
No continuity on projects

With memory:

Recalls past conversations
Knows your preferences
Tracks project history
Builds relationship over time

Goldfish → Elephant. 🐘

没有记忆的情况下：

Agent在会话之间会遗忘所有内容
重复提问，丢失上下文
项目推进没有连续性

有记忆的情况下：

可以回忆过往对话内容
了解你的使用偏好
跟踪项目历史进度
长期构建使用关联

金鱼→大象。🐘