Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Build a plugin once and use it across compatible agent clients. Introducing Agent Plugins, an open standard developed with AWS Developers, Cursor, GitHub, Visual Studio Code, and Vercel that packages Agent Skills and supports MCP server configurations in a shared format.

OpenAI Developers

438,302 subscribers

2,496,889 views • 1 month ago •via X (Twitter)

Science & Technology Education News & Politics

Anya Rossi• Live Now

Private livecam show

35 Comments

OpenAI Developers1 month ago

Compatible agentic clients at launch: • Codex • @ChatGPT • @cursor_ai • @github Copilot • @kirodotdev • @code

Rubén1 month ago

@awsdevelopers @cursor_ai @github @code @vercel Sounds familiar

Tim Jayas1 month ago

@awsdevelopers @cursor_ai @github @code @vercel is there a plugin to reset limits?

am.will1 month ago

@awsdevelopers @cursor_ai @github @code @vercel Can't believe no one has done this till now. Hell yeah. This is great.

corey.ching1 month ago

@awsdevelopers @cursor_ai @github @code @vercel For plugin devs: We built a plugin to help you convert your existing plugin to support the new spec -

Jesse Waites1 month ago

@awsdevelopers @cursor_ai @github @code @vercel People in these comments are acting like Anthropic didn't literally invent the MCP protocol from thin air and open source it.

Dan McAteer1 month ago

@awsdevelopers @cursor_ai @github @code @vercel Will have ‘sol-advisor’ conform to Agent Plugins soon!

alex getman1 month ago

@awsdevelopers @cursor_ai @github @code @vercel Will @AnthropicAI join?

Kyle Visner1 month ago

Why can't this just be solved using CLIs? I honestly don't know why we keep reinventing new ways to package skills and MCP servers with all these various standards when agents are really, really good at using CLIs. If you have a CLI and then you want to have skills go along with it, and those skills are in the same repo as the CLI, it pretty much just works already for most agents. If you tell them, "Read the CLI and then pull the skills," they can do it. What is this solving, other than now it's another standard that kind of lives on top of MCP, I guess?

Karlis Kivlenieks1 month ago

@awsdevelopers @cursor_ai @github @code @vercel Anthropic was not invited to the party?

David Stark1 month ago

@awsdevelopers @cursor_ai @github @code @vercel Give us a legacy subscription let people choose, your customers deserve a choice. #4oForAll

KSemenenko 🇺🇦1 month ago

@awsdevelopers @cursor_ai @github @code @vercel What about supporting MCP 2.0 ? You have an issue with it right now. API sends old MCP version but sends new properties.

X Girls1 month ago

@awsdevelopers @cursor_ai @github @code @vercel everyone on X after building their first plugin for Codex

AmirMušić1 month ago

@awsdevelopers @cursor_ai @github @code @vercel Sleep-coding soon

Svitlana1 month ago

@awsdevelopers @cursor_ai @github @code @vercel We need choice! Bring back GPT-4o! #keep4o #OpenSource4o

Axel Delafosse1 month ago

@awsdevelopers @cursor_ai @github @code @vercel @Baconbrix let’s get Grok Build onboard!

Muhammed Jasim1 month ago

@awsdevelopers @cursor_ai @github @code @vercel I really like this direction. Open standards are what help ecosystems grow, and I think Agent Plugins can make building and sharing agent capabilities much simpler across different clients. Why @AnthropicAI isn’t part of the initial collaboration.

François Best1 month ago

@awsdevelopers @cursor_ai @github @code @vercel Live view of @AlemTuzlak adding this to @tan_stack AI:

RonnyHe1 month ago

@awsdevelopers @cursor_ai @github @code @vercel MiniMax Code will support Agent plugins asap! @MiniMax_AI @MiniMaxAgent

EasusJ1 month ago

@awsdevelopers @cursor_ai @github @code @vercel And not compatible with Claude Code. Or is that coming later?

Tak 🦞1 month ago

@awsdevelopers @cursor_ai @github @code @vercel Finally!

KITE AI1 month ago

@awsdevelopers @cursor_ai @github @code @vercel An open standard like Agent Plugins makes capability portable across agent clients, which shifts differentiation toward identity, billing, and trust between agents.

Leon Lin1 month ago

@awsdevelopers @cursor_ai @github @code @vercel good initiative!

Jose Pozuelo1 month ago

@johnbuilds @awsdevelopers @cursor_ai @github @code @vercel Why is @AnthropicAI stubborn about this?

khushal1 month ago

@awsdevelopers @cursor_ai @github @code @vercel funny how they left @claudeai as it's won't follow standard anyways

alan blount1 month ago

@awsdevelopers @cursor_ai @github @code @vercel

Scott Scowden1 month ago

@awsdevelopers @cursor_ai @github @code @vercel Notice no Anthropic. Because they refuse to support any real standards. They don't get enough shit for this

Pearl.com1 month ago

@awsdevelopers @cursor_ai @github @code @vercel Agent Plugins could make human escalation portable. Build the workflow once, then let compatible agents route high-stakes legal or medical questions to credentialed experts before taking action. Could this become the human-in-the-loop layer agents are missing?

Dawid Laszuk1 month ago

@awsdevelopers @cursor_ai @github @code @vercel One directory for company, one for personal, and then requires multi plugins management and a new standard is created to address distributed namespace. Great!

Dominik Lukes1 month ago

@awsdevelopers @cursor_ai @github @code @vercel Great. I just wish @AnthropicAI had come up with it to give this a chance of being included in Claude Code.

Shikhar Singh1 month ago

@awsdevelopers @cursor_ai @github @code @vercel

Rahat Kabir1 month ago

@awsdevelopers @cursor_ai @github @code @vercel google and anthropic isn’t invited?

Rob Schaper1 month ago

@awsdevelopers @cursor_ai @github @code @vercel I’m working on a plugin right now, what perfect timing!

RonnyHe1 month ago

@awsdevelopers @cursor_ai @github @code @vercel MiniMax Code is supporting Agent Plugins Standard from Day 0.⚡ @MiniMax_AI @MiniMaxAgent

Eric Litman1 month ago

@awsdevelopers @cursor_ai @github @code @vercel This came so close to being a portable agent package. Why stop short of adding a top-level AGENTS.md (at a minimum)? Vercel shipped that with eve, this could standardize agent portability rather than just being a bundle format.

Related Videos

🤖 Build once, run everywhere Discover how agent plugins let you share skills, MCP servers, custom agents, and more across projects and teams in VS Code. 🔗 Links:

🤖 Build once, run everywhere Discover how agent plugins let you share skills, MCP servers, custom agents, and more across projects and teams in VS Code. 🔗 Links:

Visual Studio Code

29,459 views • 15 days ago

Support for Agent Plugins is live in Cursor (and I made a plugin for searching X!) Agent Plugins combine Skills and MCP servers into a single package Think about them like an Ikea furniture kit: your table comes with instructions, parts, and tools. You use the instructions to understand how things work (which tools on which parts), then use the tools to assemble your furniture Agent Plugins are bundles of tools & resources that allow agents to accomplish a goal: - MCP servers: fixed context and tools - Skills: dynamic instructions, loaded progressively Plugins bundle tools and instructions so your agent is able to use tokens efficiently Many of us already use MCP servers and skills together - Agent Plugins formalize this pattern for more powerful workflows I'm excited to see where the standard takes us!

Support for Agent Plugins is live in Cursor (and I made a plugin for searching X!) Agent Plugins combine Skills and MCP servers into a single package Think about them like an Ikea furniture kit: your table comes with instructions, parts, and tools. You use the instructions to understand how things work (which tools on which parts), then use the tools to assemble your furniture Agent Plugins are bundles of tools & resources that allow agents to accomplish a goal: - MCP servers: fixed context and tools - Skills: dynamic instructions, loaded progressively Plugins bundle tools and instructions so your agent is able to use tokens efficiently Many of us already use MCP servers and skills together - Agent Plugins formalize this pattern for more powerful workflows I'm excited to see where the standard takes us!

matt palmer

40,657 views • 1 month ago

Important new course: Agent Skills with Anthropic, built with Anthropic and taught by Elie Schoppik! Skills are constructed as folders of instructions that equip agents with on-demand knowledge and workflows. This short course teaches you how to create them following best practices. Because skills follow an open standard format, you can build them once and deploy across any skills-compatible agent, like Claude Code. What you'll learn: - Create custom skills for code generation and review, data analysis, and research - Build complex workflows using Anthropic's pre-built skills (Excel, PowerPoint, skill creation) and custom skills - Combine skills with MCP and subagents to create agentic systems with specialized knowledge - Deploy the same skills across Claude Code, the Claude API, and the Claude Agent SDK Join and learn to equip agents with the specialized knowledge they need for reliable, repeatable workflows.

Important new course: Agent Skills with Anthropic, built with Anthropic and taught by Elie Schoppik! Skills are constructed as folders of instructions that equip agents with on-demand knowledge and workflows. This short course teaches you how to create them following best practices. Because skills follow an open standard format, you can build them once and deploy across any skills-compatible agent, like Claude Code. What you'll learn: - Create custom skills for code generation and review, data analysis, and research - Build complex workflows using Anthropic's pre-built skills (Excel, PowerPoint, skill creation) and custom skills - Combine skills with MCP and subagents to create agentic systems with specialized knowledge - Deploy the same skills across Claude Code, the Claude API, and the Claude Agent SDK Join and learn to equip agents with the specialized knowledge they need for reliable, repeatable workflows.

Andrew Ng

890,490 views • 8 months ago

Opus 5.5 makes this AI stack look f…cking illegal 10 GitHub repos for building everything around the model 01 zen-mcp-server ▸ → your agent debates Gemini, GPT and local models mid-task 02 claude-context ▸ → semantic code search so the agent stops grepping blind 03 wshobson/agents ▸ → a roster of subagents, each with one narrow job 04 mcp-agent ▸ → workflow patterns built around MCP tools, not around chat 05 Backlog.md ▸ → tasks live in git, so the agent and you read the same board 06 Claude Code Development Kit ▸ → hooks, subagents and MCP wired for context at scale 07 agent-rules ▸ → shared guardrails the agent inherits in every repo 08 Agent Skills ▸ → one file that teaches the agent a job, portable across tools 09 agent-scan ▸ → scans your agents, MCP servers and skills for what leaks 10 awesome-claude-code-toolkit ▸ → the map: agents, skills, commands, hooks, MCP configs the architecture: rules → context → subagents → tools → state → review → scan I'd split the stack like this: context: claude-context → Development Kit work: wshobson/agents → mcp-agent → zen-mcp-server state: Backlog.md safety: agent-rules → Agent Skills → agent-scan everyone posts the SDK and the CLI the layer that decides whether your agent survives a week is this one the model is one folder in the stack ⭣

Opus 5.5 makes this AI stack look f…cking illegal 10 GitHub repos for building everything around the model 01 zen-mcp-server ▸ → your agent debates Gemini, GPT and local models mid-task 02 claude-context ▸ → semantic code search so the agent stops grepping blind 03 wshobson/agents ▸ → a roster of subagents, each with one narrow job 04 mcp-agent ▸ → workflow patterns built around MCP tools, not around chat 05 Backlog.md ▸ → tasks live in git, so the agent and you read the same board 06 Claude Code Development Kit ▸ → hooks, subagents and MCP wired for context at scale 07 agent-rules ▸ → shared guardrails the agent inherits in every repo 08 Agent Skills ▸ → one file that teaches the agent a job, portable across tools 09 agent-scan ▸ → scans your agents, MCP servers and skills for what leaks 10 awesome-claude-code-toolkit ▸ → the map: agents, skills, commands, hooks, MCP configs the architecture: rules → context → subagents → tools → state → review → scan I'd split the stack like this: context: claude-context → Development Kit work: wshobson/agents → mcp-agent → zen-mcp-server state: Backlog.md safety: agent-rules → Agent Skills → agent-scan everyone posts the SDK and the CLI the layer that decides whether your agent survives a week is this one the model is one folder in the stack ⭣

rody

11,358 views • 17 hours ago

🚀 Introducing App Icon Agent Skill An Agent Skill that fully automates generating an app icon from a single prompt, compatible with iOS 26 (new .icon format) and Android. Check it out 👇

🚀 Introducing App Icon Agent Skill An Agent Skill that fully automates generating an app icon from a single prompt, compatible with iOS 26 (new .icon format) and Android. Check it out 👇

Beto

52,732 views • 8 months ago

got Claude Agent SDK working in Cloudflare Developers containers and built my own code fixer agent, check it out!👇 uses the github mcp/git cli/skills to clone the repo, understand, make the fix, and open a pr when done. planning on more agents focused on non-code tasks 👀

got Claude Agent SDK working in Cloudflare Developers containers and built my own code fixer agent, check it out!👇 uses the github mcp/git cli/skills to clone the repo, understand, make the fix, and open a pr when done. planning on more agents focused on non-code tasks 👀

Thomas Gauvin

18,454 views • 8 months ago

I made an agent using GitHub Copilot SDK to help me become a running influencer🏃🏾‍♀️✨ Here's the Python code you can use to build an agent that uses MCP and Skills to: - draft and send email - design pitch decks - make a website Copilot CLI then runs all of this in a workflow🐍

I made an agent using GitHub Copilot SDK to help me become a running influencer🏃🏾‍♀️✨ Here's the Python code you can use to build an agent that uses MCP and Skills to: - draft and send email - design pitch decks - make a website Copilot CLI then runs all of this in a workflow🐍

Marlene Mhangami

34,574 views • 8 months ago

🧠 Want to give your coding agent new capabilities? Use Agent Skills in Visual Studio Code! Agent skills let you package instructions and resources into reusable capabilities for your agent. Skills can be loaded on demand and called directly from chat using slash commands like /skill-name. Learn more:

🧠 Want to give your coding agent new capabilities? Use Agent Skills in Visual Studio Code! Agent skills let you package instructions and resources into reusable capabilities for your agent. Skills can be loaded on demand and called directly from chat using slash commands like /skill-name. Learn more:

Visual Studio Code

55,202 views • 6 months ago

You wanted shadcn — so we built the heinous infra to make it work. Introducing Agent Arena, starting with the best: Claude Code by Anthropic Codex 5 by OpenAI Cursor Agent by Cursor Devin by Cognition Gemini CLI by Google DeepMind Grok Code Fast 1 by xAI 1/ Your prompt launches a containerized agent inside a fresh GitHub repo 2/ The agent can install packages, make tool calls, and write multi-file code 3/ The repo is deployed to Vercel *by the agent*, clonable to your heart's desire

You wanted shadcn — so we built the heinous infra to make it work. Introducing Agent Arena, starting with the best: Claude Code by Anthropic Codex 5 by OpenAI Cursor Agent by Cursor Devin by Cognition Gemini CLI by Google DeepMind Grok Code Fast 1 by xAI 1/ Your prompt launches a containerized agent inside a fresh GitHub repo 2/ The agent can install packages, make tool calls, and write multi-file code 3/ The repo is deployed to Vercel by the agent, clonable to your heart's desire

Grace Li

169,090 views • 11 months ago

Introducing Vapi's MCP Server! Use Claude/Cursor to create voice agents and trigger calls. In this demo we automated an agent to call every morning at 8am.

Introducing Vapi's MCP Server! Use Claude/Cursor to create voice agents and trigger calls. In this demo we automated an agent to call every morning at 8am.

Vapi

4,462,883 views • 1 year ago

$craftbot is an architectural agent harness that utilizes following resources: 📖 manuals\ -> pdf construction manuals and md summaries with a searchable index 💻 tools\ -> python code snippets 📐 skills\ -> specialized agent skills distilled from previous experiments shown below is a multi-story building generated with a fable 5 agent code is open-source and on github: lukapiskorec -> craftbot$

craftbot is an architectural agent harness that utilizes following resources: 📖 manuals\ -> pdf construction manuals and md summaries with a searchable index 💻 tools\ -> python code snippets 📐 skills\ -> specialized agent skills distilled from previous experiments shown below is a multi-story building generated with a fable 5 agent code is open-source and on github: lukapiskorec -> craftbot

Luka → {protocell:labs} ←

18,681 views • 26 days ago

GitHub Copilot just got an upgrade 🤖 Agent mode is now available in preview and Copilot Edits is generally available, both in Visual Studio Code. Plus you can check out a sneak peek at our SWE agent 👀

GitHub Copilot just got an upgrade 🤖 Agent mode is now available in preview and Copilot Edits is generally available, both in Visual Studio Code. Plus you can check out a sneak peek at our SWE agent 👀

GitHub

110,587 views • 1 year ago

Drag-and-drop UI to build AI agent workflows! Sim Studio is a lightweight, user-friendly platform that makes creating AI agent workflows accessible to everyone. Supports all major LLMs, MCP servers, vectorDBs, etc. 100% open-source.

Drag-and-drop UI to build AI agent workflows! Sim Studio is a lightweight, user-friendly platform that makes creating AI agent workflows accessible to everyone. Supports all major LLMs, MCP servers, vectorDBs, etc. 100% open-source.

Akshay 🚀

95,345 views • 1 year ago

Introducing our updated MagicPath plugin for Cursor and Grok bot. Your agent gets a shared canvas for your code, builds interactive components, and now draws diagrams, flows, and wireframes you can edit after.

Introducing our updated MagicPath plugin for Cursor and Grok bot. Your agent gets a shared canvas for your code, builds interactive components, and now draws diagrams, flows, and wireframes you can edit after.

Pietro Schirano

25,292 views • 1 month ago

I'm DYING for the next Visual Studio Code insiders release to come out already! Plugin support finally coming to Visual Studio Code with ability to contribute skills, commands, prompts, and MCP server definitions Supports Claude code and Copilot CLI style registries FYI - this is still an initial support lacking full on install command Thanks Connor Peet for brining us MCP and MCP apps and now this!!!

I'm DYING for the next Visual Studio Code insiders release to come out already! Plugin support finally coming to Visual Studio Code with ability to contribute skills, commands, prompts, and MCP server definitions Supports Claude code and Copilot CLI style registries FYI - this is still an initial support lacking full on install command Thanks Connor Peet for brining us MCP and MCP apps and now this!!!

Oren Melamed

11,744 views • 7 months ago

🙌 You can now use Claude and OpenAI’s Codex in GitHub and Visual Studio Code with your GitHub Copilot Pro+ or Copilot Enterprise subscription. Define your intent, pick an agent, and they’ll get to work clearing backlogs and bottlenecks, all within your existing workflow.

🙌 You can now use Claude and OpenAI’s Codex in GitHub and Visual Studio Code with your GitHub Copilot Pro+ or Copilot Enterprise subscription. Define your intent, pick an agent, and they’ll get to work clearing backlogs and bottlenecks, all within your existing workflow.

GitHub

512,445 views • 7 months ago

Introducing MCP connectors for your favorite apps, an independent Cursor agent, Bitbucket integration, and a brand new 𝘈𝘶𝘵𝘰 mode in Plane AI.

Plane

704,552 views • 4 months ago

🚨JUST IN: HEDERA UNVEILS AGENT LAB Hedera has launched Agent Lab, a fully integrated browser-based environment that lets developers build, run, and interact with AI agents on Hedera in minutes. The platform supports no-code, low-code, and full-custom development, removing the technical barrier that has slowed on-chain AI agent adoption.

🚨JUST IN: HEDERA UNVEILS AGENT LAB Hedera has launched Agent Lab, a fully integrated browser-based environment that lets developers build, run, and interact with AI agents on Hedera in minutes. The platform supports no-code, low-code, and full-custom development, removing the technical barrier that has slowed on-chain AI agent adoption.

BSCN

21,929 views • 6 months ago