· via TechCrunch
Figma adds 20 MCP partners while co-founder Dylan Field calls the protocol imperfect
Figma announced 20 new MCP partners at its first Config conference in India, even as co-founder Dylan Field said the protocol favors frontier labs and carries abuse risks.

Figma announced 20 new Model Context Protocol (MCP) partners at its first Config conference in India, according to TechCrunch, even as the design company's co-founder publicly aired doubts about the protocol itself.
The new partners include Amp, ClickUp, Dia, Junie by JetBrains, Linear, Meta Muse, Mistral, Notion, OpenCode and Sarvam Agents. MCP, created by Anthropic, defines a standard way for AI agents to plug into external tools and data sources, and the expansion means agents working across those services can increasingly reach and act on Figma data.
The backdrop: a disagreement over openness
The announcement landed shortly after a public back-and-forth about how open Figma really is. According to TechCrunch, one of MCP's co-creators had complained that Figma did not support the open-source tool Pi. In a late-September post on X, Field responded by pointing a finger back at Anthropic, noting that it does not expose artifacts or other cloud resources created via the protocol.
"I find it interesting that Anthropic (creators of MCP standard) do not expose artifacts / other cloud resources created via MCP," Field wrote, adding that "the debate is framed as open vs closed" while "issues of asymmetric context sharing are ignored." He later said Figma plans to support Pi.
A flawed but useful standard
At a media briefing during the conference, TechCrunch asked Field about the debate. His answer balanced enthusiasm for opening Figma's ecosystem with clear reservations about MCP as a piece of technology.
"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 argued that the protocol "is advantaged more to frontier labs" — the large AI developers — while acknowledging that users simply want a standard way to use their Figma data with agents, and that he wants them to have it.
Why Figma is holding back
Field, who also serves as Figma's CEO, explained the company's reluctance to open up fully in terms of risk. According to TechCrunch, he said the more data Figma exposes through MCP, the greater the chance of abuse. He added that while the vast majority of users have no malicious intent, the company still has to plan for those who might.
Why it matters
Field's criticism lands at a moment when MCP is becoming the default connector between AI agents and the software people use every day. When the leader of a major platform adopts a standard while calling it flawed and tilted toward frontier labs, it suggests MCP's dominance is pragmatic rather than enthusiastic — companies integrate because the ecosystem expects it, not because they consider the design settled.
The "asymmetric context sharing" complaint also raises a governance question that extends well past Figma. If the protocol's creator does not expose its own cloud resources through it, other platforms are effectively being asked to be more open than the party writing the rules.
Finally, the abuse rationale is likely to become a familiar one. As agents gain deeper access to design files, project trackers and documents, every company holding that data faces the same calculation between openness and exposure. Figma's answer — broaden support, but not without limits, and not without saying so publicly — may be a preview of how the rest of the industry navigates MCP in the months ahead.
- #figma
- #mcp
- #ai-agents
- #model-context-protocol
- #design-tools