mcpField
S
Spacery

Spacery

Remote
spacery.devMedia & Design2Updated 1w ago

About Spacery

Spacery is an MCP server and agent skill for publishing 3D captures as shareable virtual spaces and creating guided tours or scavenger hunts. It is for people managing spatial captures and building interactive visits.

Key features

  1. 01Publish 3D and 360 captures
  2. 02Create tours and scavenger hunts
  3. 03Draft stops with visual effects
  4. 04Manage visibility and sharing
  5. 05Upload captures and track processing

Use cases

  1. 01Publish a laser scan as a shareable space
  2. 02Create a scavenger hunt in a museum
  3. 03Add 3D models and sound to a temple tour

MCP support

Offers local stdio and hosted Streamable HTTP MCP servers. Its tools create spaces, upload captures, manage tours, and share links.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://app.spacery.dev/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Packages
mcpbspacery.mcpb0.2.3

Tools

Details

From the project README
Spacery

Spacery MCP server and agent skill

Spacery hosts 3D captures as virtual spaces with guided tours, shared with a link. This repository holds the Spacery connector (an MCP server) and agent skill, so your agent can publish captures for you and build tours and scavenger hunts in them.

  • Publish E57 and other laser scans, Matterport exports, 3D Gaussian splats, 360 photos and video, meshes and photo sets as spaces, straight from your computer.
  • Build guided tours and scavenger hunts in your own spaces or Spacery's public ones (museums, temples, tombs, gardens). Spacery's tour agent writes the stops and places 3D models, effects, sound, looks (line drawing, blueprint, film noir, night vision) and skies (a sunset, a storm, the Milky Way).
  • Your account links through a code you approve in your browser. Payment goes through Stripe Checkout in your browser. The connector never sees a password or a card.

Live demos: an observatory from a Matterport E57 and a garden as 3D Gaussian splats. Free to try. Hosting from $5 a month (pricing).

Two connectors

Local connectorHosted connector
Runson your computer, over stdioon Spacery, over Streamable HTTP
Addressnpx -y https://spacery.dev/agents/spacery-0.2.3.tgzhttps://app.spacery.dev/mcp
Reads files on your computeryes, and uploads them in the background with resumeno, it takes files the agent holds and gives you a page to drop the rest on
Use it inClaude Code, Claude Desktop, Codex, Grok Build, Antigravity, Gemini CLI, Cursorclaude.ai, ChatGPT, grok.com, Meta Muse and other apps that take a connector URL

The local connector is one readable JavaScript file with no dependencies. It needs Node.js 18 or later.

Add it

Claude Code, as a plugin with the connector and the skill.

claude plugin marketplace add lukehollis/spacery
claude plugin install spacery@spacery

Or the connector alone.

claude mcp add --scope user spacery -- npx -y https://spacery.dev/agents/spacery-0.2.3.tgz

Claude Desktop. Download spacery.mcpb and open it.

Codex.

codex mcp add spacery -- npx -y https://spacery.dev/agents/spacery-0.2.3.tgz

Gemini CLI.

gemini extensions install https://github.com/lukehollis/spacery

Grok Build.

grok mcp add spacery -- npx -y https://spacery.dev/agents/spacery-0.2.3.tgz

Antigravity.

agy mcp add spacery -- npx -y https://spacery.dev/agents/spacery-0.2.3.tgz

Cursor and any other agent that reads an mcpServers config.

{
  "mcpServers": {
    "spacery": { "command": "npx", "args": ["-y", "https://spacery.dev/agents/spacery-0.2.3.tgz"] }
  }
}

claude.ai, ChatGPT, grok.com, Meta Muse and other apps in the cloud. Add a custom connector with the URL https://app.spacery.dev/mcp.

The skill

skills/spacery/SKILL.md teaches an agent the publishing and tour steps. Add it with

npx skills add lukehollis/spacery

or save it as spacery/SKILL.md in your agent's skills folder (~/.claude/skills, ~/.agents/skills for Codex, ~/.grok/skills, ~/.gemini/config/skills for Antigravity).

Tools

ToolWhat it does
check_filesLooks at capture files and folders on this computer (local connector)
link_accountLinks the agent to your Spacery account through a code you approve in the browser
list_plansShows the ways to pay for hosting
create_spaceCreates a space and opens Stripe Checkout when hosting needs payment
wait_for_paymentWaits for Checkout to finish
upload_filesUploads a capture in the background and submits it for processing
finish_uploadSubmits files the agent uploaded itself (hosted connector)
space_statusShows upload and processing progress
set_visibilityMakes a ready space public or private
find_tour_spacesFinds your spaces and public spaces to build a tour or scavenger hunt on
create_tourStarts a guided tour or scavenger hunt on a space
draft_tourAsks Spacery's tour agent to write the tour, with models, effects, sound, looks and skies
wait_for_tourWaits for a draft and summarizes the tour
get_tour, save_tourShows a tour with everything it can use, and saves one edited by hand
search_modelsSearches the library of ready-made 3D models
upload_model, upload_skyAdds your own .glb model or 360 sky image to a tour (local connector)
list_tours, share_tourLists your tours, and shares one by link or makes it private again
unlink_accountUnlinks the agent

Try it

Ask your agent

  • "Publish the scan in ~/Captures/studio as a Spacery space."
  • "Make a scavenger hunt for five artifacts in a Spacery museum, ending in a line drawing."
  • "Model a bronze lantern in Blender and put it by the altar in my temple tour."

More

MIT licensed.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.