Most questions about a fleet of BLE tags and trackers are simple. Where is this asset? When was it last seen? Why did the data stop arriving? Until now, answering them meant opening a dashboard, filtering a list, or writing a script against an API.
Today we are opening the Nodle IoT Dashboard to AI agents through the Model Context Protocol (MCP), the open standard that lets AI assistants connect to outside tools and data. Connect Claude, ChatGPT, Gemini, Grok or any agent that supports remote MCP servers, sign in with your Nodle account, and ask your fleet directly.
With read access, your agent can:
List your fleets with their status and device count
List your devices with their fleet and last detection time
Look up a device with its last location and signal strength
Pull the recent detections of a device, with time and location
Check whether a fleet identifier is valid, and its registration
Review a fleet’s webhook status and its last 24 hours of deliveries
With write access, which you choose at sign-in, it can also:
Rename fleets and devices so names match your asset register
Turn a fleet’s existing webhook on or off
Send a signed test event to a fleet’s webhook
“Which devices in the Rotterdam fleet have not been detected since yesterday?”
“Where was tag DOCK-0042 last seen, and how strong was the signal?”
“Show me the recent detections for this tracker.”
“Is this fleet identifier valid and registered?”
“Rename these twelve devices to match the asset IDs in this list.”
“Our webhook went quiet. Check the last 24 hours of deliveries and send a test event.”
The agent picks the right tool, queries the Nodle IoT Dashboard, and answers in plain language. Because it runs inside the agent your team already uses, fleet data sits next to the rest of the work: an incident note, a status update for operations, a reconciliation against a spreadsheet.
OAuth sign-in. Every connection is authorized through a Nodle account.
Permissions you choose. Read access covers your organization, fleets, devices, detections and webhooks. Write access is optional and selected at sign-in.
Scoped and revocable. Connections are specific to a team, can be revoked at any time, and expire after 30 days.
Admin control. Administrators decide who in the organization can connect.
Limited exposure. Webhook details show the host only, and test events are capped at 5 per fleet per minute.
Server URL: https://dashboard.nodle.com/api/mcp (remote MCP over streamable HTTP, with OAuth)
Claude (web, desktop, mobile): go to Customize › Connectors, add a custom connector with the server URL, and sign in with Nodle. Team and Enterprise plans can add it once for the whole organization.
Claude Code: run
claude mcp add --transport http nodle https://dashboard.nodle.com/api/mcpChatGPT: turn on developer mode in settings, create an MCP app with OAuth, and sign in with Nodle.
Codex: add the server URL to
~/.codex/config.toml, then authenticate from the terminal.Gemini, Grok and other agents: add the same URL as a custom remote MCP server or connector, then sign in with Nodle.
Full setup guide: dashboard.nodle.com/docs/mcp
An enterprise should not have to choose its AI vendor based on what its IoT provider supports. MCP is an open standard, so your Nodle fleet follows your team’s choice of agent, today and as that choice changes.
Connect your agent at dashboard.nodle.com/docs/mcp

