mem0-memory

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Mem0 Memory Management

Mem0 内存管理

Persist and retrieve semantic memories across Claude sessions.

跨Claude会话持久化和检索语义记忆。

Memory Scopes

记忆范围

Organize memories by scope for efficient retrieval:

Scope	Purpose	Examples
`project-decisions`	Architecture and design decisions	"Use PostgreSQL with pgvector for RAG"
`project-patterns`	Code patterns and conventions	"Components use kebab-case filenames"
`project-continuity`	Session handoff context	"Working on auth refactor, PR #123 pending"

按范围组织记忆以实现高效检索：

范围	用途	示例
`project-decisions`	架构与设计决策	"Use PostgreSQL with pgvector for RAG"
`project-patterns`	代码模式与约定	"Components use kebab-case filenames"
`project-continuity`	会话交接上下文	"Working on auth refactor, PR #123 pending"

Project Isolation

项目隔离

Memories are isolated by project name extracted from

CLAUDE_PROJECT_DIR

Project name:
```
basename($CLAUDE_PROJECT_DIR)
```
(sanitized to lowercase, dashes)
Format:
```
{project-name}-{scope}
```

Edge Case: If two different repositories have the same directory name, they will share the same

user_id

scope. To avoid this:

Use unique directory names for each project
Or use
```
MEM0_ORG_ID
```
environment variable for additional namespace

Example:

```
/Users/alice/my-app
```
→
```
my-app-decisions
```
✅
```
/Users/bob/my-app
```
→
```
my-app-decisions
```
⚠️ (collision if same mem0.ai project)

With

MEM0_ORG_ID=acme

/Users/alice/my-app

→

acme-my-app-decisions

✅

记忆通过从

CLAUDE_PROJECT_DIR

中提取的项目名称进行隔离：

项目名称：
```
basename($CLAUDE_PROJECT_DIR)
```
（转换为小写，用短横线分隔）
格式：
```
{project-name}-{scope}
```

**边缘情况：**如果两个不同的仓库具有相同的目录名称，它们将共享相同的

user_id

范围。为避免这种情况：

为每个项目使用唯一的目录名称
或使用
```
MEM0_ORG_ID
```
环境变量来添加额外的命名空间

示例：

```
/Users/alice/my-app
```
→
```
my-app-decisions
```
✅
```
/Users/bob/my-app
```
→
```
my-app-decisions
```
⚠️（如果是同一个mem0.ai项目，会发生冲突）

配置

MEM0_ORG_ID=acme

后：

/Users/alice/my-app

→

acme-my-app-decisions

✅

Memory Categories

记忆分类

Memories are automatically categorized based on content. Available categories:

Category	Keywords	Use Case
`pagination`	pagination, cursor, offset	API pagination patterns
`security`	security, vulnerability, OWASP	Security patterns and vulnerabilities
`authentication`	auth, JWT, OAuth, token	Authentication patterns
`testing`	test, pytest, jest, coverage	Testing strategies
`deployment`	deploy, CI/CD, Docker, Kubernetes	Deployment patterns
`observability`	monitoring, logging, tracing, metrics	Observability patterns
`performance`	performance, cache, optimize	Performance optimization
`ai-ml`	LLM, RAG, embedding, LangChain	AI/ML patterns
`data-pipeline`	ETL, streaming, batch processing	Data pipeline patterns
`database`	database, SQL, PostgreSQL, schema	Database patterns
`api`	API, endpoint, REST, GraphQL	API design patterns
`frontend`	React, component, UI, CSS	Frontend patterns
`architecture`	architecture, design, system	Architecture patterns
`pattern`	pattern, convention, style	General patterns
`blocker`	blocked, issue, bug	Blockers and issues
`constraint`	must, cannot, required	Constraints
`decision`	chose, decided, selected	Decisions (default)

记忆会根据内容自动分类。可用分类如下：

分类	关键词	使用场景
`pagination`	pagination, cursor, offset	API分页模式
`security`	security, vulnerability, OWASP	安全模式与漏洞
`authentication`	auth, JWT, OAuth, token	认证模式
`testing`	test, pytest, jest, coverage	测试策略
`deployment`	deploy, CI/CD, Docker, Kubernetes	部署模式
`observability`	monitoring, logging, tracing, metrics	可观测性模式
`performance`	performance, cache, optimize	性能优化
`ai-ml`	LLM, RAG, embedding, LangChain	AI/ML模式
`data-pipeline`	ETL, streaming, batch processing	数据流水线模式
`database`	database, SQL, PostgreSQL, schema	数据库模式
`api`	API, endpoint, REST, GraphQL	API设计模式
`frontend`	React, component, UI, CSS	前端模式
`architecture`	architecture, design, system	架构模式
`pattern`	pattern, convention, style	通用模式
`blocker`	blocked, issue, bug	阻塞问题与故障
`constraint`	must, cannot, required	约束条件
`decision`	chose, decided, selected	决策（默认分类）

Cross-Tool Memory

跨工具记忆

Memories include

source_tool

metadata to support cross-tool memory sharing:

```
source_tool: "orchestkit-claude"
```
- Memories from Claude Code
```
source_tool: "orchestkit-cursor"
```
- Memories from Cursor (future)

Query memories by tool:

bash

undefined

记忆包含

source_tool

元数据，以支持跨工具记忆共享：

```
source_tool: "orchestkit-claude"
```
- 来自Claude Code的记忆
```
source_tool: "orchestkit-cursor"
```
- 来自Cursor的记忆（未来支持）

按工具查询记忆：

bash

undefined

Query Claude Code memories

查询Claude Code的记忆

filters={"AND": [{"metadata.source_tool": "orchestkit-claude"}]}

Query all memories (any tool)

查询所有记忆（任意工具）

filters={"AND": [{"user_id": "my-project-decisions"}]}

undefined

filters={"AND": [{"user_id": "my-project-decisions"}]}

undefined

Setup

安装配置

Install mem0 Python SDK:

bash

undefined

安装mem0 Python SDK：

bash

undefined

Install the mem0ai package and dependencies

安装mem0ai包及依赖

pip install mem0ai python-dotenv

Or install from requirements file

或从requirements文件安装

pip install -r skills/mem0-memory/scripts/requirements.txt


**Optional - Install mem0-skill-lib package (recommended for development):**

```bash

pip install -r skills/mem0-memory/scripts/requirements.txt


**可选 - 安装mem0-skill-lib包（开发推荐）：**

```bash

Install as editable package for proper type checking

以可编辑模式安装，以便进行正确的类型检查

pip install -e skills/mem0-memory/scripts/


**Note:** Scripts work in both modes:
- **Standalone mode** (default): Scripts dynamically add `lib/` to sys.path. Type checkers require `# type: ignore` comments for these dynamic imports.
- **Installed mode**: If `mem0-skill-lib` is installed, scripts import from the installed package without type ignore comments.

**Set environment variables:**

**Option 1: Using `.env` file (Recommended)**

Create a `.env` file in your project root:

```bash

pip install -e skills/mem0-memory/scripts/


**注意：**脚本支持两种运行模式：
- **独立模式**（默认）：脚本动态将`lib/`添加到sys.path。类型检查器需要为这些动态导入添加`# type: ignore`注释。
- **已安装模式**：如果安装了`mem0-skill-lib`，脚本会从已安装的包导入，无需添加类型忽略注释。

**设置环境变量：**

**选项1：使用.env文件（推荐）**

在项目根目录创建.env文件：

```bash

Copy the example file

复制示例文件

cp .env.example .env

Edit .env and add your API key

编辑.env并添加你的API密钥

MEM0_API_KEY=sk-your-api-key-here MEM0_ORG_ID=org_... # Optional (for organization-level scoping) MEM0_PROJECT_ID=proj_... # Optional (Pro feature) MEM0_WEBHOOK_URL=https://your-domain.com/webhook/mem0 # Optional


The scripts automatically load from `.env` if it exists.

**Option 2: Shell environment variables**

```bash
export MEM0_API_KEY="sk-..."
export MEM0_ORG_ID="org_..."      # Optional (for organization-level scoping)
export MEM0_PROJECT_ID="proj_..." # Optional (Pro feature)

Verify installation:

bash

python3 -c "from mem0 import MemoryClient; print('✓ mem0ai installed successfully')"

MEM0_API_KEY=sk-your-api-key-here MEM0_ORG_ID=org_... # 可选（用于组织级范围划分） MEM0_PROJECT_ID=proj_... # 可选（专业版功能） MEM0_WEBHOOK_URL=https://your-domain.com/webhook/mem0 # 可选


脚本会自动加载.env文件（如果存在）。

**选项2：Shell环境变量**

```bash
export MEM0_API_KEY="sk-..."
export MEM0_ORG_ID="org_..."      # 可选（用于组织级范围划分）
export MEM0_PROJECT_ID="proj_..." # 可选（专业版功能）

验证安装：

bash

python3 -c "from mem0 import MemoryClient; print('✓ mem0ai安装成功')"

Core Operations

核心操作

Adding Memories

添加记忆

Execute the script via Bash tool:

bash

!bash skills/mem0-memory/scripts/crud/add-memory.py \
  --text "Decided to use FastAPI over Flask for async support" \
  --user-id "project-decisions" \
  --metadata '{"scope":"project-decisions","category":"backend","date":"2026-01-12"}' \
  --enable-graph

Best practices for adding:

Be specific and actionable
Include rationale ("because...")
Add scope and category metadata
Timestamp important decisions

通过Bash工具执行脚本：

bash

!bash skills/mem0-memory/scripts/crud/add-memory.py \
  --text "Decided to use FastAPI over Flask for async support" \
  --user-id "project-decisions" \
  --metadata '{"scope":"project-decisions","category":"backend","date":"2026-01-12"}' \
  --enable-graph

添加记忆的最佳实践：

内容要具体且可执行
包含理由（"因为..."）
添加范围和分类元数据
为重要决策添加时间戳

Searching Memories

搜索记忆

bash

!bash skills/mem0-memory/scripts/crud/search-memories.py \
  --query "authentication approach" \
  --user-id "project-decisions" \
  --limit 5 \
  --enable-graph

Search tips:

Use natural language queries
Search by topic, not exact phrases
Combine with scope filters when available
Enable graph (
```
--enable-graph
```
) to get relationship information in results

Graph Relationships in Search Results:

When

--enable-graph

is enabled, search results include:

```
relations
```
array with relationship information
```
related_via
```
field showing how results are connected
```
relationship_summary
```
with relation types found

bash

!bash skills/mem0-memory/scripts/crud/search-memories.py \
  --query "authentication approach" \
  --user-id "project-decisions" \
  --limit 5 \
  --enable-graph

搜索技巧：

使用自然语言查询
按主题搜索，而非精确短语
有可用范围时结合范围过滤器
启用图功能（
```
--enable-graph
```
）以在结果中获取关系信息

搜索结果中的图关系：

启用

--enable-graph

后，搜索结果会包含：

```
relations
```
数组，包含关系信息
```
related_via
```
字段，显示结果间的关联方式
```
relationship_summary
```
，包含已发现的关系类型

Graph Relationship Queries

图关系查询

Get Related Memories:

Query memories related to a given memory via graph traversal:

bash

!bash skills/mem0-memory/scripts/graph/get-related-memories.py \
  --memory-id "mem_abc123" \
  --depth 2 \
  --relation-type "recommends"

Traverse Graph:

Multi-hop graph traversal for complex relationship queries:

bash

!bash skills/mem0-memory/scripts/graph/traverse-graph.py \
  --memory-id "mem_abc123" \
  --depth 2 \
  --relation-type "recommends"

Example Use Cases:

Multi-hop queries:

"What did database-engineer recommend about pagination?"
→ Traverses: database-engineer → recommends → cursor-pagination
→ Returns related memories with relationship context

Context expansion:

Find a memory about "authentication"
→ Get related memories via graph (depth 2)
→ Discover related decisions, patterns, and recommendations

Relationship filtering:

--relation-type "recommends"  # Only follow "recommends" relationships
--relation-type "uses"         # Only follow "uses" relationships

获取相关记忆：

通过图遍历查询与指定记忆相关的记忆：

bash

!bash skills/mem0-memory/scripts/graph/get-related-memories.py \
  --memory-id "mem_abc123" \
  --depth 2 \
  --relation-type "recommends"

遍历图：

用于复杂关系查询的多跳图遍历：

bash

!bash skills/mem0-memory/scripts/graph/traverse-graph.py \
  --memory-id "mem_abc123" \
  --depth 2 \
  --relation-type "recommends"

示例用例：

多跳查询：

"数据库工程师关于分页的建议是什么？"
→ 遍历路径：database-engineer → recommends → cursor-pagination
→ 返回带有关系上下文的相关记忆

上下文扩展：

查找关于"authentication"的记忆
→ 通过图获取相关记忆（深度2）
→ 发现相关的决策、模式和建议

关系过滤：

--relation-type "recommends"  # 仅遵循"推荐"关系
--relation-type "uses"         # 仅遵循"使用"关系

Listing Memories

列出记忆

bash

!bash skills/mem0-memory/scripts/crud/get-memories.py \
  --user-id "project-orchestkit" \
  --filters '{"limit":100}'

bash

!bash skills/mem0-memory/scripts/crud/get-memories.py \
  --user-id "project-orchestkit" \
  --filters '{"limit":100}'

Getting Single Memory

获取单个记忆

bash

!bash skills/mem0-memory/scripts/crud/get-memory.py \
  --memory-id "mem_abc123"

bash

!bash skills/mem0-memory/scripts/crud/get-memory.py \
  --memory-id "mem_abc123"

Updating Memories

更新记忆

bash

!bash skills/mem0-memory/scripts/crud/update-memory.py \
  --memory-id "mem_abc123" \
  --text "Updated decision text" \
  --metadata '{"updated":true}'

bash

!bash skills/mem0-memory/scripts/crud/update-memory.py \
  --memory-id "mem_abc123" \
  --text "Updated decision text" \
  --metadata '{"updated":true}'

Deleting Memories

删除记忆

bash

!bash skills/mem0-memory/scripts/crud/delete-memory.py \
  --memory-id "mem_abc123"

When to delete:

Outdated decisions that were reversed
Incorrect information
Duplicate or redundant entries

bash

!bash skills/mem0-memory/scripts/crud/delete-memory.py \
  --memory-id "mem_abc123"

删除场景：

已被推翻的过时决策
错误信息
重复或冗余条目

What to Remember

记忆存储建议

Good candidates:

Architecture decisions with rationale
API contracts and interfaces
Naming conventions adopted
Technical debt acknowledged
Blockers and their resolutions
User preferences and style

Avoid storing:

Temporary debugging context
Large code blocks (use Git)
Secrets or credentials
Highly volatile information

适合存储的内容：

带有理由的架构决策
API契约与接口
已采用的命名约定
已确认的技术债务
阻塞问题及其解决方案
用户偏好与风格

避免存储的内容：

临时调试上下文
大段代码（使用Git存储）
密钥或凭证
高度易变的信息

Memory Patterns

记忆模式

Decision Memory

决策记忆

"Decision: Use cursor-based pagination for all list endpoints.
Rationale: Better performance for large datasets, consistent UX.
Date: 2026-01-12. Scope: API design."

"决策：所有列表端点使用基于游标分页。
理由：大数据集下性能更优，用户体验一致。
日期：2026-01-12。范围：API设计。"

Pattern Memory

模式记忆

"Pattern: All React components export default function.
Convention: Use named exports only for utilities.
Applies to: frontend/src/components/**"

"模式：所有React组件导出默认函数。
约定：仅工具函数使用命名导出。
适用范围：frontend/src/components/**"

Continuity Memory

会话衔接记忆

"Session handoff: Completed hybrid search implementation.
Next steps: Add metadata boosting, write integration tests.
PR #456 ready for review. Blocked on: DB migration approval."

"会话交接：已完成混合搜索实现。
下一步：添加元数据增强，编写集成测试。
PR #456待审核。阻塞项：数据库迁移审批。"

Advanced Operations (Pro Features)

高级操作（专业版功能）

Batch Operations

批量操作

Batch update up to 1000 memories:

bash

!bash skills/mem0-memory/scripts/batch/batch-update.py \
  --memories '[{"memory_id":"mem_123","text":"updated"},{"memory_id":"mem_456","metadata":{"updated":true}}]'

Batch delete:

bash

!bash skills/mem0-memory/scripts/batch/batch-delete.py \
  --memory-ids '["mem_123","mem_456","mem_789"]'

批量更新最多1000条记忆：

bash

!bash skills/mem0-memory/scripts/batch/batch-update.py \
  --memories '[{"memory_id":"mem_123","text":"updated"},{"memory_id":"mem_456","metadata":{"updated":true}}]'

批量删除：

bash

!bash skills/mem0-memory/scripts/batch/batch-delete.py \
  --memory-ids '["mem_123","mem_456","mem_789"]'

Memory History (Audit Trail)

记忆历史（审计追踪）

bash

!bash skills/mem0-memory/scripts/utils/memory-history.py \
  --memory-id "mem_abc123"

bash

!bash skills/mem0-memory/scripts/utils/memory-history.py \
  --memory-id "mem_abc123"

Exports (Data Portability)

导出（数据可移植性）

Create export:

bash

!bash skills/mem0-memory/scripts/export/export-memories.py \
  --filters '{"user_id":"project-decisions"}' \
  --schema '{"format":"json"}'

Retrieve export:

bash

!bash skills/mem0-memory/scripts/export/get-export.py \
  --user-id "project-decisions"

创建导出任务：

bash

!bash skills/mem0-memory/scripts/export/export-memories.py \
  --filters '{"user_id":"project-decisions"}' \
  --schema '{"format":"json"}'

获取导出数据：

bash

!bash skills/mem0-memory/scripts/export/get-export.py \
  --user-id "project-decisions"

Analytics

分析

Get memory statistics:

bash

!bash skills/mem0-memory/scripts/utils/memory-summary.py \
  --filters '{"user_id":"project-decisions"}'

List all users:

bash

!bash skills/mem0-memory/scripts/utils/get-users.py

获取记忆统计数据：

bash

!bash skills/mem0-memory/scripts/utils/memory-summary.py \
  --filters '{"user_id":"project-decisions"}'

列出所有用户：

bash

!bash skills/mem0-memory/scripts/utils/get-users.py

Webhooks (Automation)

Webhook（自动化）

bash

!bash skills/mem0-memory/scripts/webhooks/create-webhook.py \
  --url "https://example.com/webhook" \
  --name "Memory Webhook" \
  --event-types '["memory.created","memory.updated"]'

bash

!bash skills/mem0-memory/scripts/webhooks/create-webhook.py \
  --url "https://example.com/webhook" \
  --name "Memory Webhook" \
  --event-types '["memory.created","memory.updated"]'

Integration with OrchestKit

与OrchestKit集成

Use memories to maintain context across plugin sessions:

bash

undefined

使用记忆在插件会话间维护上下文：

bash

undefined

At session start - recall project context

会话开始时 - 召回项目上下文

!bash skills/mem0-memory/scripts/crud/search-memories.py
--query "current sprint priorities"
--user-id "project-continuity"

During work - persist decisions

工作过程中 - 持久化决策

!bash skills/mem0-memory/scripts/crud/add-memory.py
--text "Implemented feature using approach because reason"
--user-id "project-decisions"
--metadata '{"scope":"project-decisions"}'

At session end - save continuity

会话结束时 - 保存衔接信息

!bash skills/mem0-memory/scripts/crud/add-memory.py
--text "Session end: summary. Next: next_steps"
--user-id "project-continuity"
--metadata '{"scope":"project-continuity"}'

undefined

!bash skills/mem0-memory/scripts/crud/add-memory.py
--text "Session end: summary. Next: next_steps"
--user-id "project-continuity"
--metadata '{"scope":"project-continuity"}'

undefined

Scripts Available

可用脚本

All scripts are located in

skills/mem0-memory/scripts/

Core Scripts:

```
add-memory.py
```
- Store new memory
```
search-memories.py
```
- Semantic search
```
get-memories.py
```
- List all memories (with filters)
```
get-memory.py
```
- Get single memory by ID
```
update-memory.py
```
- Update memory content/metadata
```
delete-memory.py
```
- Remove memory

Advanced Scripts (Pro Features):

```
batch-update.py
```
- Bulk update up to 1000 memories
```
batch-delete.py
```
- Bulk delete multiple memories
```
memory-history.py
```
- Get audit trail for a memory
```
export-memories.py
```
- Create structured export
```
get-export.py
```
- Retrieve export data
```
memory-summary.py
```
- Get statistics/analytics
```
get-events.py
```
- Track async operations
```
get-users.py
```
- List all users (analytics)
```
create-webhook.py
```
- Setup webhooks for automation

Note: Mem0 uses CLI scripts via Bash (not MCP). Scripts provide full control, versioning, and access to all 30+ API methods. The primary memory layer is the knowledge graph (

mcp__memory__*

); mem0 CLI scripts serve as an optional cloud enhancement for semantic search across sessions.

所有脚本位于

skills/mem0-memory/scripts/

目录下：

核心脚本：

```
add-memory.py
```
- 存储新记忆
```
search-memories.py
```
- 语义搜索
```
get-memories.py
```
- 列出所有记忆（支持过滤）
```
get-memory.py
```
- 通过ID获取单个记忆
```
update-memory.py
```
- 更新记忆内容/元数据
```
delete-memory.py
```
- 删除记忆

高级脚本（专业版功能）：

```
batch-update.py
```
- 批量更新最多1000条记忆
```
batch-delete.py
```
- 批量删除多条记忆
```
memory-history.py
```
- 获取记忆的审计追踪
```
export-memories.py
```
- 创建结构化导出
```
get-export.py
```
- 获取导出数据
```
memory-summary.py
```
- 获取统计/分析数据
```
get-events.py
```
- 跟踪异步操作
```
get-users.py
```
- 列出所有用户（分析用）
```
create-webhook.py
```
- 配置自动化Webhook

**注意：**Mem0通过Bash使用CLI脚本（而非MCP）。脚本提供完整控制、版本管理，并可访问所有30+ API方法。核心记忆层是知识图谱（

mcp__memory__*

）；mem0 CLI脚本作为可选的云增强功能，用于跨会话的语义搜索。

Related Skills

Key Decisions

关键决策

Decision	Choice	Rationale
Memory scope	project-decisions, project-patterns, project-continuity	Clear separation of memory types
Storage format	Natural language with metadata	Semantic search works best with descriptive text
MCP integration	Mem0 MCP server	Native Claude Desktop integration
What to avoid	Secrets, large code blocks, volatile info	Keep memories clean and safe

决策	选择	理由
记忆范围	project-decisions, project-patterns, project-continuity	清晰区分记忆类型
存储格式	带元数据的自然语言	语义搜索在描述性文本上表现最佳
MCP集成	Mem0 MCP服务器	原生Claude Desktop集成
避免存储内容	密钥、大段代码、易变信息	保持记忆干净安全

Capability Details

功能详情

memory-add

Keywords: add memory, remember, store, persist, save context Solves:

How do I save information for later sessions?
Persist a decision or pattern
Store project context across sessions

关键词： add memory, remember, store, persist, save context 解决问题：

如何为后续会话保存信息？
持久化决策或模式
跨会话存储项目上下文

memory-search

Keywords: search memory, recall, find, retrieve, what did we Solves:

How do I find previous decisions?
Recall context from past sessions
Search for specific patterns or conventions

关键词： search memory, recall, find, retrieve, what did we 解决问题：

如何查找之前的决策？
召回过往会话的上下文
搜索特定模式或约定

memory-list

Keywords: list memories, show all, get memories, view stored Solves:

How do I see all stored memories?
List project decisions
Review stored patterns

关键词： list memories, show all, get memories, view stored 解决问题：

如何查看所有已存储的记忆？
列出项目决策
回顾已存储的模式

memory-delete

Keywords: delete memory, forget, remove, clear Solves:

How do I remove outdated memories?
Delete incorrect information
Clean up duplicate entries

关键词： delete memory, forget, remove, clear 解决问题：

如何移除过时的记忆？
删除错误信息
清理重复条目

mem0-memory

Original

Translation

Mem0 Memory Management

Mem0 内存管理

Memory Scopes

记忆范围

Project Isolation

项目隔离

Memory Categories

记忆分类

Cross-Tool Memory

跨工具记忆

Query Claude Code memories

查询Claude Code的记忆

Query all memories (any tool)

查询所有记忆（任意工具）

Setup

安装配置

Install the mem0ai package and dependencies

安装mem0ai包及依赖

Or install from requirements file

或从requirements文件安装

Install as editable package for proper type checking

以可编辑模式安装，以便进行正确的类型检查

Copy the example file

复制示例文件

Edit .env and add your API key

编辑.env并添加你的API密钥

Core Operations

核心操作

Adding Memories

添加记忆

Searching Memories

搜索记忆

Graph Relationship Queries

图关系查询

Listing Memories

列出记忆

Getting Single Memory

获取单个记忆

Updating Memories

更新记忆

Deleting Memories

删除记忆

What to Remember

记忆存储建议

Memory Patterns

记忆模式

Decision Memory

决策记忆

Pattern Memory

模式记忆

Continuity Memory

会话衔接记忆

Advanced Operations (Pro Features)

高级操作（专业版功能）

Batch Operations

批量操作

Memory History (Audit Trail)

记忆历史（审计追踪）

Exports (Data Portability)

导出（数据可移植性）

Analytics

分析

Webhooks (Automation)

Webhook（自动化）

Integration with OrchestKit

与OrchestKit集成

At session start - recall project context

会话开始时 - 召回项目上下文

During work - persist decisions

工作过程中 - 持久化决策

At session end - save continuity

会话结束时 - 保存衔接信息

Scripts Available

可用脚本

Related Skills

相关技能

Key Decisions