London-based analytics firm Edited is bringing its retail dataset directly into AI workspaces, with its newest product powered by the Model Context Protocol, an open standard created by Anthropic.
On Wednesday Edited launched Edited MCP, which allows technical teams to connect the world’s largest aggregation of retail data into Claude Code, Claude Desktop and custom-built AI agents without having to export a file or switch tabs.
The company emphasized that general-purpose AI tools are only as good as the data behind them. While large language models typically lack the ability to verify market context, Edited MCP has data from 90,000 brands and over five billion stock-keeping units across home, beauty and apparel.
Anthropic compares a MCP to “a USB-C port for AI applications,” connecting AI applications to external systems. Edited said the MCP is a focused and lightweight data layer built for technical teams already building their own agents. An internal tool will now be able to pull real-time market context back in seconds, rather than stopping to search or guess.
“Our customers didn’t wait for us to build this,” said Doug Kofoid, CEO at Edited. “They were already wiring their own planning agents and copilots. Most AI projects fail for the same reason: the data underneath isn’t good enough to trust. We’ve spent twelve years solving that problem. Now retail AI teams finally have something worth building on.”
This marks Edited’s latest AI-powered offering, following the release of a conversational workspace powered by Edited’s own AI Edie last July.
Edited MCP launches first for builders, developers and technical teams building the agents and workflows their organizations will run on. The setup takes an API key configured as a request header, which is the same pattern used across professional developer tools, giving technical teams full control from day one.
Confirmed and tested support at launch includes Claude Code, Claude Desktop, VS Code, and Cursor, with technical teams also able to connect Edited MCP directly into their own custom-built agents.
“An open standard only matters if what’s behind it is worth connecting to,” said David Hartig, chief technology officer at Edited. “Anthropic built the protocol. We spent twelve years building the dataset. Put those together, and a developer gets a real answer in seconds, not a guess dressed up to sound confident.”
This reflects the direction the industry is going, where AI becomes a part of operations, may it be in the backend or customer facing operations like AI chatbots. Before this AI pivot, Edited was already sought out by retailers for its data analytics.
