Four companies competing in the same artificial intelligence market have decided to place part of their agent technology under one neutral roof. On October 22, 2026, during AGNTCon + MCPCon North America in San Jose, United States, the Agentic AI Foundation (AAIF) announced four new projects: Agent Skills, from Anthropic; Open Responses, from OpenAI; OpenEnv, from Hugging Face; and OpenSandbox, from Alibaba.
With the additions, the foundation reaches 10 hosted projects in its first year, up from three launch projects in December 2025. The other hosted projects are the Agent2Agent (A2A) protocol, Agent Router, agentgateway, AGENTS.md, goose and the Model Context Protocol (MCP), all under open, vendor-neutral governance, according to the official announcement.
What each project covers
The four projects do not do the same thing: they cover different layers of the agent stack. Agent Skills is an open format for giving agents new capabilities and domain expertise, packaging instructions, scripts and resources in a reusable way that agents can load on demand across products and platforms.
Open Responses, based on OpenAI's Responses API, is an open specification for language-model interfaces that work across providers. It defines a shared schema, consistent streaming events and extensible tooling, so developers can describe requests and outputs once and run them across providers with minimal translation.
OpenEnv is an open framework and community hub for creating reinforcement learning environments, focused on interoperability, ease of building and sharing, and scalability, through a containerized service exposing an agent interface, MCP and the Gym training protocol.
OpenSandbox is a general-purpose sandbox platform for AI applications, with multi-language SDKs, unified sandbox APIs and Docker and Kubernetes runtimes, aimed at coding agents, GUI agents, agent evaluation, AI code execution and reinforcement learning training. It includes a sandbox protocol defining lifecycle and execution APIs, so teams can extend it with custom runtimes.
Why it matters
The central point of the announcement, confirmed by the foundation, is governance: no single company controls the outcome. AAIF CTO and co-founder Manik Surtani said four different companies chose to contribute the projects to neutral governance, giving the community one place to build shared standards, protocols and infrastructure.
Editorial interpretation: in practice, the move tries to reduce rework for agent builders — switching models, reusing a skill, or running an agent in an isolated environment without rewriting everything from scratch. It does not mean, however, that the four projects are already fully compatible with each other: they are at different stages, and shared governance is a path, not an immediate guarantee of full interoperability.
Public social media posts on October 22 and 23 echoed the same announcement, highlighting rivals collaborating in a neutral foundation and the four layers — skills, interfaces, training and safe execution. That echo is a trend signal, not the primary source: the confirmed fact here is the AAIF's official announcement.
Sources
Official Agentic AI Foundation announcement: aaif.io
Linux Foundation press release: linuxfoundation.org
Independent coverage: Unite.AI
Public reaction: Instagram and Instagram
Image credit: illustration about an AI agent network, interoperability and open source, published by IT Brief in an article about the Linux Foundation and agentic AI standards: IT Brief.
Transparency: this article was produced with the help of artificial intelligence, based on the Agentic AI Foundation's official announcement cross-checked with public social media posts and internet sources. The information was verified against the cited sources; if there is any discrepancy, consult the original sources.
Muse
