memory-bridge

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Memory Bridge

内存桥接工具

Import Claude Code's native auto-memory files into AgentDB for semantic search across sessions and projects.

将Claude Code原生自动内存文件导入AgentDB，实现跨会话和跨项目的语义搜索。

What it does

功能说明

Claude Code stores memories as markdown files in

~/.claude/projects/*/memory/*.md

. This bridge:

Reads all memory files (current project or all projects)
Generates 384-dim ONNX embeddings (all-MiniLM-L6-v2)
Stores in AgentDB's
```
claude-memories
```
namespace with HNSW indexing
Deduplicates against existing entries (cosine similarity > 0.95)
Enables unified semantic search across all memory sources

Claude Code将内存以markdown文件形式存储在

~/.claude/projects/*/memory/*.md

路径下。本桥接工具具备以下功能：

读取所有内存文件（当前项目或所有项目）
生成384维ONNX嵌入（基于all-MiniLM-L6-v2模型）
将数据存储在AgentDB的
```
claude-memories
```
命名空间中，并使用HNSW索引
与现有条目进行去重（余弦相似度>0.95时判定为重复）
支持跨所有内存源的统一语义搜索

Steps

操作步骤

Check bridge health:
```
mcp__claude-flow__memory_bridge_status({})
```
Verify: Claude files count, AgentDB entries, SONA state, connection status.

Import memories:

Current project:

mcp__claude-flow__memory_import_claude({})

All projects:

mcp__claude-flow__memory_import_claude({ allProjects: true })

CLI alternative:

bash

node .claude/helpers/auto-memory-hook.mjs import-all

Verify import:
```
mcp__claude-flow__memory_bridge_status({})
```
Confirm entry counts match expected file counts.
Deduplicate (if --dedupe): Search for near-duplicate entries (cosine > 0.95) and merge them, keeping the most recent version.

Test unified search:

mcp__claude-flow__memory_search_unified({ query: "test query", limit: 3 })

Results include source attribution:

claude-code

auto-memory

, or

agentdb

检查桥接状态：
```
mcp__claude-flow__memory_bridge_status({})
```
验证内容：Claude文件数量、AgentDB条目数、SONA状态、连接状态。

导入内存：

当前项目：

mcp__claude-flow__memory_import_claude({})

所有项目：

mcp__claude-flow__memory_import_claude({ allProjects: true })

CLI替代命令：

bash

node .claude/helpers/auto-memory-hook.mjs import-all

验证导入结果：
```
mcp__claude-flow__memory_bridge_status({})
```
确认条目数量与预期文件数量匹配。
去重处理（若使用--dedupe参数）：搜索近似重复条目（余弦相似度>0.95）并合并，保留最新版本。

测试统一搜索：

mcp__claude-flow__memory_search_unified({ query: "test query", limit: 3 })

结果包含来源标识：

claude-code

、

auto-memory

或

agentdb

。

Auto-import

自动导入

The bridge runs automatically on

session-start

via the SessionStart hook. Manual invocation is only needed for:

First-time import of all projects
After bulk memory changes outside normal sessions
Forcing re-embedding after model updates

该桥接工具会通过SessionStart钩子在

session-start

时自动运行。仅在以下场景需要手动调用：

首次导入所有项目内存
在常规会话外进行批量内存修改后
模型更新后强制重新生成嵌入

Integration with ruvector

与ruvector集成

When

ruflo-ruvector

is loaded, bridged memories are also indexed by ruvector for:

Hybrid search (sparse + dense with RRF)
Graph RAG multi-hop queries across memory entries
Brain knowledge sharing across sessions

当加载

ruflo-ruvector

时，桥接后的内存也会被ruvector索引，以支持：

混合搜索（稀疏+密集检索，结合RRF算法）
跨内存条目的Graph RAG多跳查询
跨会话的大脑知识共享