At its first Config conference held in India, Figma announced that it is broadening support for the open-source Model Context Protocol (MCP), adding twenty new MCP partners. The newly announced partners include Amp, ClickUp, Dia, Junie by JetBrains, Linear, Meta Muse, Mistral, Notion, OpenCode and Sarvam Agents.
The expansion comes amid reservations voiced by Dylan Field, co-founder of Figma, about the protocol. Field has previously expressed concerns about MCP; on September 30, 2026 he posted on X that he found it notable that Anthropic — one of the creators of the MCP standard — does not expose artifacts or other cloud resources created via MCP. In that post he argued the debate is often framed as open versus closed while issues of asymmetric context sharing are overlooked.
(Field later added to his post that Figma plans to support Pi.)
At a media briefing after the conference I asked Field about the debate. He said that while Figma wants to open its ecosystem, he does not consider MCP a perfect protocol. “I’m really excited about what we’re doing in terms of opening up our ecosystem. I also think that MCP is not a perfect protocol. And I think that it is something that has tons of room for improvement,” Field said. He added that the protocol is advantaged toward frontier labs, and that he understands users who want a straightforward protocol to use their Figma data.
Figma’s CEO also discussed the trade-offs between openness and risk. He explained that exposing more data through MCP increases the likelihood of abuse. Although most users do not intend to misuse Figma data, the company must account for those potential risks when deciding how broadly to expose resources via MCP.
Why this matters
Figma’s announcement signals rapid growth in integrations between design tools and AI agents: adding twenty MCP partners is a meaningful step toward greater interoperability. At the same time, the comments from Dylan Field and Figma’s CEO underscore the tension between technical openness and the security and privacy concerns that arise when sharing contextual data across services.
Going forward, developments such as whether Anthropic makes more of its resources available through MCP and how Figma implements support for Pi will influence how responsibilities and capabilities are distributed across the emerging AI ecosystem.
