LLMemory
Enables VS Code agent mode to search and retrieve context from your LLMemory vault of AI conversations (ChatGPT, Claude, etc.) without copying and pasting.
README
LLMemory for VS Code
Connect VS Code agent mode to your LLMemory vault — search and pull context from your ChatGPT, Claude, Codex, and Claude Code conversations without copy-pasting.
What it does
Developers working across multiple AI tools (Claude Code, ChatGPT, Codex) constantly lose context when switching between them. LLMemory lets VS Code's agent mode pull prior sessions or hand off context without leaving the editor — "pull my most recent Codex session" or "what did I figure out about the OAuth worker last week?" — instead of copy-pasting.
Install
- Install this extension from the VS Code Marketplace.
- Open agent mode in Chat view (
Cmd+I/Ctrl+I). - LLMemory appears in the MCP SERVERS section of the Extensions view — enabled by default.
- On first tool call, a browser OAuth flow grants Google Drive access (
drive.filescope, limited to theLLMemory/folder).
No configuration needed. The extension registers the remote MCP server automatically.
Tools
| Tool | Description |
|---|---|
list_sources |
Show which AI providers have data in your vault |
list_recent |
List most recently updated conversations |
search_chats |
Full-text search across all imported AI conversations |
get_conversation |
Fetch one full conversation transcript by ID |
get_context_pack |
Build an agent-ready briefing for handoff |
Privacy / architecture
- Zero server-side storage — the Cloudflare Worker brokers Google OAuth and reads your Drive live; no chat content, no per-user database. Free tier.
- Your own Google Drive is the storage — written by the LLMemory browser extension / CLI.
- Read-only — the connector never writes to the vault.
- OAuth is fully automated via Dynamic Client Registration — no API keys to manage.
Requirements
- VS Code 1.101.0 or later
- An LLMemory vault (install the browser extension to start capturing chats)
Links
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。