OpenAI, Microsoft, and Cursor Unite Behind Agent Plugins to End Fragmented AI Workflows
OpenAI, Microsoft, Cursor, AWS, and Vercel launch Agent Plugins v1.0: one plugin format that works across Codex, ChatGPT, Copilot, VS Code, Cursor, and Kiro.
- Agent Plugins v1.0 is a new open standard for packaging Agent Skills and MCP server configs into one portable plugin directory that works across multiple agent clients.
- Launch clients include ChatGPT, Codex, GitHub Copilot, VS Code, Cursor, and Kiro (AWS) -- all loading the same plugin format.
- The Technical Steering Committee includes Amazon, Cursor, Microsoft, OpenAI, and Vercel, making this a genuine cross-industry standard.
- Agent Skills (originally from Anthropic) bundle instructions, scripts, and reference docs into a simple folder with a
SKILL.mdfile -- no server required. - MCP server configs can be bundled in the same plugin via
mcp.json, supporting stdio, Streamable HTTP, and legacy SSE transports. - The spec is free and open at agent-plugins.org; proposals and governance happen publicly on GitHub.
Every major coding agent has had its own plugin format. A workflow you packaged for GitHub Copilot needed rework before Cursor could use it. A tool integration wired up for Codex wouldn't load in VS Code. The result was a fragmented ecosystem where plugin authors had to maintain parallel versions of the same thing. Agent Plugins is the industry's answer to that problem.
One package, every client
Agent Plugins is an open, vendor-neutral standard for packaging reusable components into portable plugins. Its version 1.0.0 specification defines a shared format for Agent Skills and MCP servers that compatible clients can discover and load consistently. The Technical Steering Committee behind it includes core maintainers from Amazon, Cursor, Microsoft, OpenAI, and Vercel.
The launch clients are:
- ChatGPT and Codex (OpenAI)
- GitHub Copilot (Microsoft)
- VS Code (Microsoft)
- Cursor
- Kiro (AWS)
AI agent clients had developed their own plugin formats, even when plugins contained the same underlying components. Authors had to rearrange or duplicate those components for each client, so a plugin packaged for one client might need adaptation before another could use it. Agent Plugins defines the minimum shared structure so the portable parts stay portable, while each client keeps full control over installation, permissions, and UX.
What's actually inside a plugin
An Agent Plugin is just a directory. The structure is intentionally minimal:
my-plugin/
├── plugin.json # identity + spec version
├── skills/
│ └── summarize/
│ └── SKILL.md # instructions for the agent
├── mcp.json # MCP server config
└── com.example.client/ # client-specific extensions