Launch of Agentic AI Interactive Dashboards MCP Connector

We are pleased to announce the Sprucely.io MCP connector for agentic AI. Claude and any other MCP-compatible AI assistant can now turn data from a conversation or an agent run into a persistent, hosted, interactive dashboard that can be shared and updated over time, with no Sprucely.io account required for viewers.

This release introduces:

• A hosted MCP server, reachable over Streamable HTTP, that connects as a custom connector in Claude, with a single command in Claude Code, or from any other MCP-compatible client
• Eight tools covering the full dashboard lifecycle: create, list and inspect dashboards, append new data, read a dataset’s schema, control sharing, fetch embed code, and check storage usage
• Two ways to authenticate - sign in with a Sprucely.io account through OAuth 2.1, or a long-lived MCP API key for headless and scheduled agents
• Dashboards that grow over time - an agent can append fresh rows to a dashboard it created earlier on any cadence, including unattended scheduled runs, so a link shared once keeps showing the latest results

For setup instructions and examples, see the Claude MCP connector guide and the full MCP API reference. The dataset and dashboard JSON the tools read and write is the same format documented on the data format page and used throughout the service.