For AI assistants

Use DEPLOY in your AI assistant

Connect Claude (or any MCP-aware assistant) to the DEPLOY registry. Ask questions about robot deployments, AI brains, incidents, and regulations and get answers grounded in verified data instead of training memory.

Install for Claude Desktop

Download the extension. Double-click the file to install. Restart Claude.

Download .dxt (5 KB) For Claude Desktop 0.10+

Or paste the URL into any MCP client

If your client supports remote MCP servers (Claude Desktop, Cursor, Claude Code, and most newer MCP clients do), just add this URL in the app's MCP settings:

Step-by-step: Claude Desktop

Recent Claude Desktop versions (2025+) support adding remote MCP servers from the UI:

Open Claude Desktop
Settings → Developer → Edit Config (or use the .dxt download above for one-click)

If using config editor, paste:

{
  "mcpServers": {
    "deploy-registry": {
      "url": "https://registry.deploy.report/mcp"
    }
  }
}

Save, restart Claude
You should see "deploy-registry" in the tool list (small icon in the input bar)

On older Claude Desktop versions that don't support remote MCP, use the bridge config below.

Bridge config (works on every version):

{
  "mcpServers": {
    "deploy-registry": {
      "command": "npx",
      "args": ["-y", "mcp-remote@latest", "https://registry.deploy.report/mcp"]
    }
  }
}

Step-by-step: Cursor

Open Cursor
Settings → Features → MCP (or Settings → MCP depending on version)
Click "Add new MCP server"
Name: deploy-registry
If Cursor offers a "URL" type, paste: https://registry.deploy.report/mcp
Otherwise use command type: command npx, args ["-y", "mcp-remote@latest", "https://registry.deploy.report/mcp"]
Save, restart Cursor

Step-by-step: Claude Code, Continue, other MCP clients

Most MCP clients accept either a direct URL or a stdio bridge command. Use:

URL: https://registry.deploy.report/mcp
Bridge command: npx -y mcp-remote@latest https://registry.deploy.report/mcp

Refer to your client's MCP documentation for where to enter these.

Try these prompts

Once installed, open a new conversation and try:

What robots does Waymo operate in San Francisco?
Show me recent incidents involving humanoid robots
Get the DEPLOY record for Figure 02
List verified deployments of the 1X NEO

Your AI assistant calls the registry tools automatically when relevant. You don't have to invoke them manually.

What's available

The MCP server exposes 5 tools that wrap DEPLOY's verified registry:

search_entities: Search the DEPLOY registry across companies, models, deployments, brains, incidents, regulations, and locations. Returns matching entities with canonical names, registry URLs, and entity types.
get_entity: Get full details for one entity by its canonical UUID. Returns the entity record plus schema.org JSON-LD structured data. Use search_entities first to resolve a name to a UUID.
get_deployments_at_location: List verified robot deployments at a specific location (by location slug). Returns deployment records with model, operator, status, and dates.
get_deployments_for_model: List verified deployments of a specific robot model (by model slug). Returns deployments with location, operator, and dates.
get_recent_incidents: List the most recently recorded incidents. Returns headline, occurredAt, kind, and slug. Useful for surfacing the latest events in the registry.

Why this matters

AI assistants are typically limited to training-set knowledge, which goes stale fast on a domain that's evolving as quickly as physical AI. The DEPLOY registry is live, sourced, and human-reviewed. Connecting your assistant to it means answers about robots, deployments, and incidents reflect what's actually verified in the world, not what the training data caught.

DEPLOY also distinguishes verified deployments from manufacturer claims. When your assistant queries DEPLOY, that distinction surfaces in the response. Read our methodology for the full editorial framework.

Help + troubleshooting

If something doesn't work:

Confirm the server is reachable: https://registry.deploy.report/mcp in a browser should show this page
If using the bridge config and seeing "npx not found", install Node.js first
If your client doesn't support MCP at all yet, use the REST API directly
For corrections or partnership questions, see cite

For developers

Direct API access:

Endpoint: POST https://registry.deploy.report/mcp
Protocol: JSON-RPC 2.0 over HTTP. Stateless. Anonymous read-only.

Test from a terminal:

curl -X POST https://registry.deploy.report/mcp \
  -H 'Content-Type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'