Cornerstone MCP Server

Cornerstone MCP Server

Enables Claude and other MCP clients to interact with Cornerstone memory, supporting saving, searching, and managing contextual data through natural language commands.

Category
访问服务器

README

Cornerstone MCP Server

MCP server providing 17+ tools for Claude Code, Claude Desktop, and any MCP-compatible client to interact with Cornerstone memory.

Quick Start

  1. Clone and set up:

    git clone https://github.com/supramal1/cornerstoneMCP.git
    cd cornerstoneMCP
    bash setup.sh
    
  2. Edit .env with your API URL and key

  3. Add to Claude Code settings

  4. Start using it:

    • "Remember that the project deadline is March 15" → saves to memory
    • "What do we know about the Google pitch?" → recalls from memory
    • "Forget the old project deadline" → removes from memory

Deploy

Deployed to Google Cloud Run (europe-west2). See cornerstone repo for deploy scripts.

Auth

OAuth 2.1 with JWT principal-based access control. Falls back to legacy API key auth.

Configuration

Required env vars: CORNERSTONE_URL, CORNERSTONE_API_KEY, OAUTH_JWT_SECRET, MCP_PUBLIC_URL.

Optional: DISABLE_OAUTH=true skips OAuth advertisement entirely (no /.well-known/oauth-*, no WWW-Authenticate: Bearer ...). Use when this MCP is fronted by a gateway (e.g. obot) that owns user auth and proxies inbound traffic without an Authorization header. Backend calls still authenticate via CORNERSTONE_API_KEY.

Tools

Simple (start here)

Tool Purpose Example
remember Save anything to memory remember("The budget is $50K")
recall Search memory recall("client's email?")
forget Remove from memory forget("old_deadline", confirm=True)

Workspace

Tool Purpose
list_workspaces See available workspaces
switch_workspace Change active workspace
get_current_workspace See which workspace is active
set_default_workspace Set default workspace

Advanced

Tool Purpose
get_context Full context retrieval with metadata
add_fact Save a structured key-value fact
add_note Save a freeform note
search Search across all memory types
list_facts List recent facts
get_recent_sessions See recent conversations
list_threads See conversation threads
report_context_feedback Rate retrieval quality

Review (MC-2 propose-then-approve)

Tool Purpose
mc_list_proposals List agent-proposed facts awaiting human review
mc_promote_facts Promote proposed facts to current memory
mc_reject_facts Reject proposed facts (reason required, stored verbatim)

Promote/reject relay the gateway-injected X-Forwarded-User-Email header on the outbound API call so the decision is attributed to the human reviewer. Absent the header (e.g. stdio), the decision is recorded without a forwarded identity.

Most users only need remember and recall. The advanced tools give you more control when you need it.

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选