wanderlog-mcp
Wanderlog-mcp is an MCP server that lets LLM agents plan and edit Wanderlog trip itineraries — adding places, hotels, notes, and checklists to any day, managing trip dates, and searching for destinations. It turns a conversational prompt into a fully built-out travel itinerary with practical details interleaved between each stop.
README
wanderlog-mcp
An MCP server that lets Claude (or any MCP-compatible agent) view and build Wanderlog trip itineraries through conversation.
Instead of clicking through the Wanderlog UI to plan a trip, you just ask:
"Create a 14-day Japan Golden Route trip — Tokyo, Hakone, Kyoto, Nara, and Osaka."
The agent calls the tools, interleaves places and notes for each day, adds hotel blocks and checklists, and you end up with a fully populated Wanderlog trip in a few minutes.
See a real example: 14-day Japan Golden Route — built entirely by an AI agent using this MCP server.
What's New in v0.1.0
- Full itinerary building: places, notes, hotels, and checklists in a single conversation
wanderlog_search_places— find real-world places near any destination using Wanderlog's place databasewanderlog_add_note— interleave transit tips, booking info, and local advice between placeswanderlog_add_checklist— pre-trip and per-day checklists (visa, currency, timed-entry tickets)- MCP server instructions injected at startup so Claude builds complete itineraries automatically
- Startup auth probe — catches expired cookies immediately instead of failing mid-conversation
Example Prompts
"What trips do I have in Wanderlog?"
"Create a 7-day itinerary for Lisbon starting June 1 — include restaurants, day trips,
and a hotel near the waterfront."
"Add a day trip to Sintra on day 3 of my Lisbon trip."
"I'm spending 5 days in Tokyo — build me a full itinerary with museum visits, ramen spots,
and a ryokan in Shinjuku."
"Look at my Barcelona trip and add practical notes for getting between each place."
"Add a pre-trip checklist to my Paris trip — visa, currency, offline maps, travel insurance."
"Move my Rome trip back by two weeks."
"Give me the shareable link to my Kyoto itinerary."
"Remove the Colosseum from day 2 of my Rome trip."
Tools
| Tool | What it does |
|---|---|
wanderlog_list_trips |
List trips in your account |
wanderlog_get_trip |
View a full itinerary, or filter to a single day |
wanderlog_get_trip_url |
Get a shareable wanderlog.com link |
wanderlog_search_places |
Find real-world places near a trip's destination |
wanderlog_create_trip |
Create a new trip with destination + date range |
wanderlog_add_place |
Add a place to a specific day or general list |
wanderlog_add_note |
Add a note (transit tips, booking info, local advice) |
wanderlog_add_hotel |
Add a hotel booking with check-in/check-out dates |
wanderlog_add_checklist |
Add a pre-trip or per-day checklist |
wanderlog_remove_place |
Remove a place by natural-language reference |
wanderlog_update_trip_dates |
Change a trip's date range |
Prerequisites
- Node.js 22 or newer
- A Wanderlog account
- An MCP-compatible client: Claude Code, Claude Desktop, OpenAI Codex, Cursor, VS Code, or any stdio MCP host
Setup
Step 1 — Get your Wanderlog session cookie
Wanderlog doesn't have a public API, so wanderlog-mcp authenticates using your browser session cookie (connect.sid). It's valid for roughly a year and never leaves your machine.
Treat it like a password — it grants the same access you have in the Wanderlog UI.
Chrome / Edge
- Go to wanderlog.com and log in
- Press
F12to open DevTools - Click the Application tab
- In the left sidebar expand Storage → Cookies → https://wanderlog.com
- Find the row where Name is
connect.sid - Click the row, then double-click the Value cell and copy the full string — it starts with
s%3Aand is ~100 characters long
Firefox
- Go to wanderlog.com and log in
- Press
F12to open DevTools - Click the Storage tab
- In the left sidebar expand Cookies → https://wanderlog.com
- Find
connect.sidin the table, click it, and copy the Value
Why can't I use
document.cookiein the console? Wanderlog setsconnect.sidwith theHttpOnlyflag, which deliberately blocks JavaScript from reading it (XSS protection). DevTools bypasses this restriction — that's why it works and the console doesn't.
Step 2 — Configure your MCP client
Claude Code
claude mcp add wanderlog-mcp npx wanderlog-mcp \
--env WANDERLOG_COOKIE="s%3A...your value here..."
Claude Desktop
Edit claude_desktop_config.json:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json - Linux:
~/.config/Claude/claude_desktop_config.json
{
"mcpServers": {
"wanderlog": {
"command": "npx",
"args": ["wanderlog-mcp"],
"env": {
"WANDERLOG_COOKIE": "s%3A...your value here..."
}
}
}
}
Restart Claude Desktop after saving.
Cursor
Settings → MCP → Add server, or edit ~/.cursor/mcp.json:
{
"mcpServers": {
"wanderlog": {
"command": "npx",
"args": ["wanderlog-mcp"],
"env": {
"WANDERLOG_COOKIE": "s%3A...your value here..."
}
}
}
}
VS Code (GitHub Copilot)
Add to your workspace .vscode/mcp.json:
{
"servers": {
"wanderlog": {
"type": "stdio",
"command": "npx",
"args": ["wanderlog-mcp"],
"env": {
"WANDERLOG_COOKIE": "s%3A...your value here..."
}
}
}
}
OpenAI Codex
Edit ~/.codex/config.toml:
[mcp_servers.wanderlog]
command = "npx"
args = ["wanderlog-mcp"]
[mcp_servers.wanderlog.env]
WANDERLOG_COOKIE = "s%3A...your value here..."
Run /mcp inside Codex to confirm the server loaded.
Smithery (one-click install)
npx @smithery/cli install wanderlog-mcp --client claude
Step 3 — Verify
Ask your agent: "What trips do I have in Wanderlog?"
It should call wanderlog_list_trips and return your account's trips. If it fails, see Troubleshooting below.
Refreshing your cookie
The cookie lasts about a year but can die sooner if you log out of wanderlog.com, change your password, or Wanderlog revokes the session. When that happens every tool call returns:
Wanderlog session invalid or expired — Capture a fresh connect.sid cookie from wanderlog.com DevTools and update WANDERLOG_COOKIE in your MCP config.
Repeat Step 1 above, update your config, and restart your MCP client.
Troubleshooting
Server starts but list_trips returns an auth error Your cookie is expired or wrong. Re-capture it from DevTools and update your config.
npx wanderlog-mcp hangs or does nothing
The server speaks stdio MCP — it's designed to be launched by an MCP host, not run directly in a terminal. Run it through Claude Code or Claude Desktop as described above.
Tools work but the agent ignores notes/checklists
The server injects instructions into the MCP initialize response that tell the agent to interleave places and notes and add checklists. This works reliably with Claude. Other clients may vary.
Security
- The cookie is stored only in your MCP client config, never committed or logged
- wanderlog-mcp runs entirely on your machine — there's no relay server
- The startup auth probe validates your cookie without printing its value
- To revoke access: log out of wanderlog.com (invalidates all sessions), then re-capture
Contributing
Pull requests are welcome. Before submitting:
npm run build && npm run test
For changes to transport or tool code, also run:
npm run test:integration
Disclaimer
wanderlog-mcp is an unofficial third-party tool, not affiliated with or endorsed by Wanderlog. It works by calling Wanderlog's private web-client API, which may change without notice. Use at your own risk.
License
MIT — see LICENSE
Made by shaikhspeare
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。