My Next Adventure

cloud.mynextadventurePublished by mynextadventure.cloudOfficial publisherv1.0.1Updated Oct 24, 2026

Create trips, compare variants side by side, and add accommodation, transport and event options.

VerifiedStreamable HTTPWeb executableLocation & LifestyleProductivity & Workflow

Overview

AI-generated overview

Remote MCP server that lets an assistant build and edit MyNextAdventure trips: variants, destinations, flights, hotels and activities.

What you can do
  • Check the connected account and list trips or travel goals
  • Read a full trip with its variants, destinations, options and events
  • Create and update trips, variants and destinations, and reorder stops
  • Propose accommodation, transport and getting-around options, then select or edit them
  • Add, edit, toggle or delete activities, and create public read-only share links
When to use it
Useful when you plan trips in MyNextAdventure and want an assistant to draft itineraries, compare plan variants, gather hotel and flight options, and keep activities organised through the conversation.
Requirements
A remote Streamable HTTP endpoint; no local install needed. Requires a MyNextAdventure account and an API key created in the app under Settings, sent as an Authorization Bearer header or an X-API-Key header. A local stdio mode exists for subprocess clients, using the MNA_API_KEY variable.
Before you install
The API key grants whatever the account can do, so the assistant can create, edit and permanently delete trips, variants, destinations, options and events; deleting a whole variant needs an explicit confirm flag. Share links expose a trip publicly in read-only form. Some tools resolve places through Google Places and other third-party APIs, sending those lookups outside MyNextAdventure. Revoke keys in the app if exposed.

Installation

In SourceWeft

  1. Open My Next Adventure in the dashboard and add it to a workspace.
  2. Connect the required credentials. They are stored privately and never shown on public pages.
  3. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/cloud.mynextadventure%2Ftrip-planner/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config. The client completes OAuth sign-in on first connect.

{
  "mcpServers": {
    "my-next-adventure": {
      "type": "http",
      "url": "https://mcp.mynextadventure.cloud/mcp"
    }
  }
}

README

MyNextAdventure MCP server

A remote Model Context Protocol server that lets an AI assistant plan trips in MyNextAdventure — create a trip, lay out the itinerary, propose flights and hotels, add things to do, and share the result.

It runs as a stateless Cloudflare Worker and is a thin, authenticated proxy in front of the MyNextAdventure public API: it holds no trip data of its own, and forwards every call to the API under the caller's own API key.

  • Production: https://mcp.mynextadventure.cloud/mcp (Streamable HTTP)
  • Upstream API: https://api.mynextadventure.cloud (v1 public API)

Authentication

Every request needs a MyNextAdventure API key. Create one in the app under Settings → API keys, then send it as either:

  • Authorization: Bearer <api-key> — what most MCP clients send, and what mcp-remote does, or
  • X-API-Key: <api-key> — for clients that only let you set custom headers.

The worker forwards it to the API as X-API-Key. Requests without a key get a 401. Whatever the key can do, the assistant can do: keys are per-user, so the assistant only ever sees that user's trips.

Where your key lives. The server is stateless: your key is read from the headers of each request, used to make that one call to the MyNextAdventure API, and never stored — there is no session storage and no Durable Object holding it between requests. It is never logged and is not shared between requests or users. Revoke a key any time in the app under Settings → API keys.

Tools

24 tools, covering the operations conversational planning actually needs. Each one is a curated view over a MyNextAdventure API operation — see Regenerating the tools.

ToolWhat it does
whoamiWhich account the API key belongs to (handy for checking the connection).
list_tripsList the user's trips; filter by status (planning, ready, finished, cancelled).
get_tripOne trip in full — variants, destinations, options, events. Needed before any edit.
create_tripCreate a new, empty trip.
update_tripRename, set a cover photo, or move the trip through its lifecycle.
create_trip_share_linkCreate/refresh a public read-only share link.
create_variantAdd an alternative plan (dates live here: exact or flexible).
duplicate_variantCopy a variant with everything in it.
update_variantRename a variant, change its dates or notes.
select_variantMark a variant as the chosen plan.
add_destinationAdd a stop to a variant, in itinerary order.
update_destinationChange a stop's place name, dates or notes.
reorder_destinationsReorder the stops in a variant.
add_accommodation_optionPropose a place to stay at a destination.
add_transport_optionPropose a way of getting to a destination.
add_getting_around_optionPropose local mobility (public transport, rental car, …).
update_destination_optionEdit an existing option of any of the three kinds.
select_destination_optionSelect — or, with no optionId, deselect — an option.
add_eventAdd an activity (museum, concert, restaurant, tour) to a variant.
update_eventEdit an activity.
toggle_eventFlip an activity between committed and "just an idea".
delete_trip_itemPermanently delete a variant, destination, option or event. Deleting a whole variant additionally requires confirm: true.
list_goalsList the user's travel goals / bucket list.
add_goalAdd a goal from a name or a URL.

The server also ships MCP instructions describing the data model (trip → variants → destinations → options, with events on the variant) and the usual order of operations, so a client does not have to guess the workflow.

Not exposed yet, though the API supports them: collaborator access management and invites, voting, collections, invite tokens, goal↔trip linking, API key management, and trip deletion.

Tool annotations

Every tool publishes a human-readable title (also as annotations.title; set in scripts/tool-manifest.ts, required by the Claude connectors directory) and all four MCP ToolAnnotations hints on tools/list, so clients can tell at a glance what is safe to run unattended:

HintTrue for
readOnlyHintThe 4 lookups: whoami, list_trips, get_trip, list_goals. The other 20 write.
destructiveHintdelete_trip_item only. Everything else adds records or edits fields in place.
idempotentHintThe update_* / select_* tools, reorder_destinations, create_trip_share_link (the API returns the existing token) and delete_trip_item. False for every add_*/create_* — a second call adds a second record — and for toggle_event, which flips a flag.
openWorldHintThe 8 tools whose request can reach a third-party API: the three read tools (Kinde profiles, Google Places, currency conversion) plus add_accommodation_option, add_transport_option, add_event, update_destination_option and update_event, which resolve a googlePlaceId through Google Places when one is supplied. The rest only touch MyNextAdventure storage.

The spec's defaults are the pessimistic ones (destructiveHint and openWorldHint both default to true), so the manifest requires all four to be stated per tool rather than inherited. test/annotations.test.ts holds the expected classification tool by tool and cross-checks readOnlyHint against the HTTP verb each tool actually issues.

Connecting

Claude Desktop

Settings → Developer → Edit Config, then add:

json
{  "mcpServers": {    "mynextadventure": {      "command": "npx",      "args": [        "mcp-remote",        "https://mcp.mynextadventure.cloud/mcp",        "--header",        "Authorization: Bearer ${MNA_API_KEY}"      ],      "env": { "MNA_API_KEY": "your-api-key" }    }  }}

Restart Claude Desktop; the tools appear under the connector.

claude.ai / Claude Code

Add a custom connector pointing at https://mcp.mynextadventure.cloud/mcp and set the Authorization: Bearer <api-key> header. From Claude Code:

bash
claude mcp add --transport http mynextadventure https://mcp.mynextadventure.cloud/mcp \  --header "Authorization: Bearer <api-key>"

ChatGPT

Settings → Connectors → Add custom connector, MCP server URL https://mcp.mynextadventure.cloud/mcp, authentication "custom header" with X-API-Key: <api-key> (or Authorization: Bearer <api-key>).

Any other MCP client

Streamable HTTP at /mcp. GET / returns a small JSON health document with the endpoint and tool count.

Local stdio

The same tools over stdio, for clients that launch a subprocess, with the key in MNA_API_KEY. Glama's build check runs this behind mcp-proxy.

bash
npm ciMNA_API_KEY=<api-key> npm run start:stdio

Development

bash
npm installnpm run dev              # wrangler dev on http://localhost:8787npm test                 # vitest: request building, schemas, generator freshnessnpm run type-check       # tsc --noEmitnpm run generate:tools   # regenerate src/generated/tools.ts from the OpenAPI specnpx tsx scripts/smoke.ts # drive the MCP protocol against localhost:8787

scripts/smoke.ts performs a real handshake (initialize → tools/list → one tools/call), reading the API key from $MNA_API_KEY or ~/.config/mna/credentials. It refuses to call mutating tools unless --force is passed. To exercise write tools without touching real data, point the worker at a local mock:

bash
npx wrangler dev --var MNA_API_BASE_URL:http://localhost:8899

Regenerating the tools

Tool schemas are generated, not hand-written:

spec/openapi-v1.json          (the API's OpenAPI snapshot)        +scripts/tool-manifest.ts      (which operations become tools; names, descriptions, annotations)        ↓  scripts/generator.tssrc/generated/tools.ts        (committed; zod input schemas, annotations + request builders)

To add or change a tool, edit scripts/tool-manifest.ts and run npm run generate:tools. Never edit src/generated/tools.ts by hand — a test fails if it drifts from the spec.

A manifest entry names one API operation, or groups several body-less operations behind a discriminator argument (delete_trip_item's target, select_destination_option's action) to keep the tool count down. Path params can be renamed (id → tripId), query params can be narrowed to enums, and descriptions can be overridden — everything else, including the request-body schemas, comes from the spec.

Each entry must also carry annotations (all four MCP hints; see Tool annotations). TypeScript enforces this, so a new tool cannot be added without classifying it, and test/annotations.test.ts checks the result against an expected table.

Deployment

npm run deploy (wrangler) — production only, so it is gated on a maintainer; CI never deploys this worker.

Development

npm cinpm run type-checknpm test        # includes a drift test: src/generated/tools.ts must match the generator outputnpm run dev     # wrangler dev

Tools are generated from the vendored OpenAPI snapshot in spec/openapi-v1.json (canonical live spec: https://api.mynextadventure.cloud/api/v1/openapi.json) via scripts/generate-tools.ts. Regenerate after refreshing the snapshot.

This repo is the deployable source for mcp.mynextadventure.cloud. Development also happens in the My Next Adventure monorepo and is synced here on release. Issues and PRs welcome.

License

MIT

Source: README.md at commit 343f2ef

Tools

0
This server asks you to sign in, so SourceWeft could not read its tools without an account and has not indexed them. What a connected account sees can depend on that account and its permissions.

Version history

2
  1. v1.0.1LatestOct 5, 2026
  2. v1.0.0Sep 16, 2026

Score

85% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts