galaxy-mcp
Live MCP server for the Galaxy vault. Lets Claude (mobile/web) and other chat interfaces query your Obsidian vault in real time via GitHub.
README
Galaxy MCP Server
Live MCP server for the Galaxy vault. Lets Claude (mobile/web) and other chat interfaces query your Obsidian vault in real time via GitHub.
Setup — do this once
Step 1: Create a private GitHub repo for the vault
- Go to github.com → New repository
- Name it
galaxy-vault, set it to Private - Do NOT initialize with README
Step 2: Install Obsidian Git and push the vault
- In Obsidian: Settings → Community Plugins → Browse → search "Obsidian Git" → Install → Enable
- Open the Obsidian Git settings:
- Auto pull interval:
5(minutes) - Auto push interval:
5(minutes) - Commit message:
vault: auto-sync {{date}}
- Auto pull interval:
- Run command palette → "Obsidian Git: Initialize a new repo"
- Run command palette → "Obsidian Git: Open source control view"
- Stage all → Commit → set remote to your new private repo URL → Push
After this, every edit to the vault auto-pushes to GitHub within 5 minutes.
Step 3: Create a GitHub Personal Access Token
- github.com → Settings → Developer Settings → Personal access tokens → Fine-grained tokens
- Name:
galaxy-mcp - Repository access: select
galaxy-vaultonly - Permissions: Contents → Read-only
- Copy the token
Step 4: Deploy to Railway
- Go to railway.app → New Project → Deploy from GitHub repo → select this
galaxy-mcprepo - Add environment variables (Settings → Variables):
GITHUB_TOKEN= (the token from step 3)
GITHUB_OWNER= (your GitHub username)
GITHUB_REPO= galaxy-vault
API_KEYS= (two random keys, comma-separated — run `openssl rand -hex 32` twice)
- Railway will auto-deploy. Copy your public URL (e.g.
https://galaxy-mcp-production.up.railway.app)
Connecting to Claude.ai (mobile + web)
- Open claude.ai → Settings → Integrations → Add MCP Server
- URL:
https://your-railway-url.up.railway.app/mcp - Auth: Bearer token → paste Priya's API key
- For Vinod: repeat with his API key
Connecting to ChatGPT (Custom GPT)
ChatGPT doesn't speak MCP natively. Add it as a Custom Action:
- Create a Custom GPT → Configure → Add Action
- Import schema from:
https://your-railway-url.up.railway.app/openapi.json - Auth: API Key → Bearer → paste API key
(The server serves an OpenAPI schema automatically at /openapi.json)
Available tools
| Tool | What it does |
|---|---|
get_schedule |
Returns current-week.md + summer schedule |
get_context |
Loads context files (family, self, home, arya, raya, courtney) |
read_file |
Read any vault file by path |
list_files |
List all markdown files |
search_vault |
Search across all files for a keyword |
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。