Reference
The standards, and who supports them
Composable parts need shared interfaces. This page lists the open standards for each layer of an agent system, and a dated matrix of which agents an official source lists for each standard.
The standards layer
The open standards that make agents composable
Composable parts need shared interfaces. The Agentic AI Foundation (AAIF) names five layers of an “open agentic stack”, and one of its projects covers each layer. The Linux Foundation hosts the AAIF.
The table adds Agent Skills, two MCP extensions, and projects for discovery and conformance testing.
Source: AAIF: A2A joins AAIF. State on .
| Layer | Standard | Steward and status | Source |
|---|---|---|---|
| The five AAIF layers | |||
| Instructions and context | AGENTS.md | OpenAI released it in August 2025. OpenAI gave it to the AAIF when the AAIF formed on 9 December 2025. | AAIF formation announcement; AAIF: A2A joins AAIF |
| Agent runtime | goose | Block gave it to the AAIF on 9 December 2025. | AAIF formation announcement; AAIF: A2A joins AAIF |
| Agent to tool connectivity | Model Context Protocol (MCP) | Anthropic released it as open source in November 2024. Anthropic gave it to the AAIF on 9 December 2025. | AAIF formation announcement; AAIF: A2A joins AAIF |
| Traffic mediation and control | agentgateway | An AAIF project. It sits between agent systems and their infrastructure, and it manages routing, policy and observability. | AAIF: A2A joins AAIF |
| Agent to agent interoperability | Agent2Agent (A2A) | Google launched it in April 2025 and gave it to the Linux Foundation. It joined the AAIF as a hosted project on 17 August 2026. More than 150 organizations back it. | AAIF: A2A joins AAIF |
| Skills and MCP extensions | |||
| Packaged procedures | Agent Skills | A skill is a folder with a SKILL.md file. Anthropic first developed the format and released it as an open standard. | Agent Skills overview |
| Skills inside MCP | Skills over MCP extensionio.modelcontextprotocol/skills | The specification has Final status. The Skills Over MCP Working Group maintains it. Hosts and software development kits are still adding support. | MCP Skills extension |
| Interactive interfaces inside MCP | MCP Apps extensionio.modelcontextprotocol/ui | An extension to the core MCP specification. MCP servers can return interactive interfaces, such as forms and dashboards, that show in the chat. Host support varies by client. | MCP Apps overview |
| Discovery | |||
| Find MCP servers | Official MCP Registry | In preview. The official central metadata repository for publicly accessible MCP servers. | The MCP Registry |
| Find MCP servers | GitHub MCP Registry | GitHub launched it on 16 September 2025 as a place to discover MCP servers. | The GitHub Blog |
| Find A2A agents and MCP servers | AGNTCY directories | Cisco released AGNTCY as open source in March 2025. The Linux Foundation welcomed the project on 29 July 2025. Its directories make A2A agents and MCP servers discoverable. | Linux Foundation press release |
| Conformance testing | |||
| Test MCP clients and servers | MCP Conformance Test Framework | Tests MCP client and server implementations against the specification. | modelcontextprotocol/ |
| Test A2A implementations | A2A Technology Compatibility Kit (TCK) | A compatibility test suite for A2A implementations on three transports. | a2aproject/ |
Compatibility matrix
Which agents support which standards
The matrix shows which agents each official source lists. The matrix covers six agents: Claude Code, ChatGPT and Codex, Cursor, GitHub Copilot with Visual Studio Code (VS Code), Gemini CLI (command-line interface) and goose.
Each column names its source and the date of the last check. In the A2A column, each cell cites the documentation of that client.
- Listed: An official source names the client as supporting the standard.
- Listed, partial: The source names the client and labels its support “Partial”.
- Not listed: The source that we checked does not name the client. This makes no claim either way.
| Client | Agent SkillsAgent Skills Client ShowcaseChecked | AGENTS.mdagents.md compatibility listChecked | MCP AppsMCP Extension Support MatrixChecked | Skills over MCPMCP Extension Support MatrixChecked | A2AClient documentationChecked |
|---|---|---|---|---|---|
| Claude Code | Listed | ListedNot on the agents.md list.Source: Claude Code Docs: AGENTS.md | Not listedThe matrix lists Claude (web) and Claude Desktop. | Not listed | Not listedSearched: Claude Code Docs |
| ChatGPT and Codex | ListedAs “ChatGPT & Codex” | ListedAs “Codex” | ListedAs “ChatGPT” | Listed, partialAs “ChatGPT” | Not listedSearched: ChatGPT Learn |
| Cursor | Listed | Listed | Listed | Not listed | Not listedSearched: Cursor Docs |
| GitHub Copilot and VS Code | ListedAs “GitHub Copilot” and “VS Code” | ListedAs “Coding agent from GitHub Copilot” and “VS Code” | ListedAs “VS Code GitHub Copilot” | Not listed | Not listedSearched: GitHub Copilot docsChecked |
| Gemini CLI | Listed | Listed | Not listed | Not listed | ListedRemote subagents.Source: Gemini CLI docs: Remote Subagents |
| goose | Listed | Listed | Listed | Not listed | Not listedSearched: goose docsChecked |
Column sources
- Agent Skills: Agent Skills: Client Showcase. Checked .
- AGENTS.md: AGENTS.md: One AGENTS.md works across many agents. Checked .
- MCP Apps and Skills over MCP: Model Context Protocol: Extension Support Matrix. Checked . The page says that the community maintains this list.
- A2A: The documentation of each client. Checked .
Cells with their own source
- AGENTS.md, Claude Code: Claude Code Docs: How Claude remembers your project: AGENTS.md. Checked . The page says: “Claude Code can read AGENTS.md as your project instructions”.
- A2A, Gemini CLI: Gemini CLI documentation: Remote Subagents. Checked . The page says: “Gemini CLI supports connecting to remote subagents using the Agent-to-Agent (A2A) protocol.”
- A2A, the other five clients: we searched the documentation of each client for “A2A”, “Agent2Agent” and “agent-to-agent”. We found no statement of A2A support. The method for each cell is in
matrix.json.- Claude Code: Claude Code Docs
- ChatGPT and Codex: ChatGPT Learn, OpenAI developer docs
- Cursor: Cursor Docs
- GitHub Copilot and VS Code: GitHub Copilot docs, VS Code docs
- goose: goose docs
Notes
- Names in quotation marks show how the source writes the client name.
- The same data, with a source and a check date for each cell:
matrix.json.
Related terms
Composability, orchestration and chaining
People often use these three words as if they mean the same thing. They do not.
Composability is a design property. Orchestration and chaining are runtime patterns: they describe how the parts run.
Composability
You build the system from independent units. You can swap or recombine the units.
Source: Atlan
Composability makes both runtime patterns possible. Atlan gives an example: units with clean boundaries “can be chained today and orchestrated tomorrow without a rewrite.”Source: Atlan
Method
How this page stays true
30 matrix cells: 18 listed, 12 not listed.
We check each source by hand.
Each matrix cell uses one official source: the list that the standard publishes, or the documentation of the client.
Each cell records its source and the date of its last check. When a source changes, we change the cell and its date.