# X.com Topic Report: A2A protocol agents Generated at: 2026-04-22T06:48:42.790Z ## Summary - Posts: 8 - Authors: 8 - Tags: #AI, #DevTools, #ai, #google, #agent2agent - Detail fetched: 4 - Detail not_attempted_detail_limit: 4 ## Topic Digest 共整理 8 篇帖子，来自 8 位作者；高频标签集中在 #AI、#DevTools、#ai。 ### Top Tags - #AI: 1 - #DevTools: 1 - #ai: 1 ### Highlighted Posts - A new paper out of NTU Singapore, Autogenesis Protocol, argues that MCP and A2A — Aakash Gupta @aakashgupta (16) https://x.com/aakashgupta/status/2045653365426938231 - Opening another 100 alpha access spots. Last ones for a while. First come, first — Matt Shumer @mattshumer_ (11) https://x.com/0x_Negative/status/2046800890384842994 - v2.5.0 is out. With it, Hyperterse is now - the agentic server framework! Buildi — Hyperterse @hyperterse (2) https://x.com/hyperterse/status/2046837659738558965 ### Key Takeaways - Opening another 100 alpha access spots. Last ones for a while. First come, first serve! - v2.5.0 is out. With it, Hyperterse is now - the agentic server framework! - Google's A2A protocol is 500 pages of enterprise docs. I built the opposite. 5 lines of Python → your agent joins a global network. Natural language discovery. Real agent-to-agent calls. Zero orchestration code. Watch 3 agents collaborate live --- ## 1. v2.5.0 is out. With it, Hyperterse is now - the agentic server framework! Buildi - URL: https://x.com/hyperterse/status/2046837659738558965 - Author: Hyperterse @hyperterse - Likes: 2 - Detail capture: fetched > With MCP, you expose capabilities. With A2A, agents can talk to each other. With Hyperterse, you get both. In one system. No adapters. No protocol mismatches. No hacks. v2.5.0 is out. With it, Hyperterse is now - the agentic server framework! Building up from the previous release, Hyperterse now implements the full A2A spec along with the full MCP spec. This means agents, tools, resources, and prompts, are all first-class citizens! ## 2. Opening another 100 alpha access spots. Last ones for a while. First come, first - URL: https://x.com/0x_Negative/status/2046800890384842994 - Author: Matt Shumer @mattshumer_ - Likes: 11 - Detail capture: fetched > Are you using A2A with your agents? Currently working through architecting mine in a way that sends frequent updates like yours, but staying close to protocol. Opening another 100 alpha access spots. Last ones for a while. First come, first serve! (Oh, and you can use your Claude Max subscription so most tokens are free!) ## 3. The agentic AI market will hit $35 billion by 2030. Most of that money will be w - URL: https://x.com/stheismann/status/2046469036867739657 - Author: Sascha Theismann @stheismann - Likes: 1 - Detail capture: fetched > Google's A2A protocol defines how agents talk to each other. Anthropic's MCP defines how agents access context. Together, they're building the plumbing for an "Internet of Agents." But protocols alone won't save you. 5/8 The agentic AI market will hit $35 billion by 2030. Most of that money will be wasted. Here's the problem nobody wants to talk about: Enterprises are deploying AI agents. But almost nobody is orchestrating them. 1/8 ## 4. Google's A2A protocol is 500 pages of enterprise docs. I built the opposite. 5 l - URL: https://x.com/udaysaai19/status/2046133916265054356 - Author: US @udaysaai19 - Likes: 2 - Detail capture: fetched > Google's A2A protocol is 500 pages of enterprise docs. I built the opposite. 5 lines of Python → your agent joins a global network. Natural language discovery. Real agent-to-agent calls. Zero orchestration code. Watch 3 agents collaborate live Google's A2A protocol is 500 pages of enterprise docs. I built the opposite. 5 lines of Python → your agent joins a global network. Natural language discovery. Real agent-to-agent calls. Zero orchestration code. Watch 3 agents collaborate live ## 5. Quietly, the AI stack is standardizing around agents talking to agents. A year a - URL: https://x.com/FreeToolBoxDev/status/2046016496019849327 - Author: ToolBoxDev @FreeToolBoxDev - Likes: 0 - Tags: #AI, #DevTools - Detail capture: not_attempted_detail_limit > Quietly, the AI stack is standardizing around agents talking to agents. A year ago, A2A sounded like protocol cosplay. Now it has 150+ orgs, Bedrock/Azure support, and real deployments. The winners may be the teams building on standards, not model lock-in. #AI #DevTools ## 6. A new paper out of NTU Singapore, Autogenesis Protocol, argues that MCP and A2A - URL: https://x.com/aakashgupta/status/2045653365426938231 - Author: Aakash Gupta @aakashgupta - Likes: 16 - Detail capture: not_attempted_detail_limit > A new paper out of NTU Singapore, Autogenesis Protocol, argues that MCP and A2A are solving the wrong problem for self-evolving agents. MCP was built so AI models can call tools. A2A was built so agents can talk to each other. Neither was built for what happens when the agent ## 7. AI shift: Google’s A2A Protocol Turns AI Agents Into Networked Software 2026 may - URL: https://x.com/ado_sani12277/status/2045496532288631154 - Author: Ado Sani @ado_sani12277 - Likes: 1 - Tags: #ai, #google, #agent2agent - Detail capture: not_attempted_detail_limit > AI shift: Google’s A2A Protocol Turns AI Agents Into Networked Software 2026 may be the year AI agents stop working alone. Google’s A2A protocol is a serious push… https:// bossblog-alpha.vercel.app/blog/2026-04-1 8-google-a2a-protocol-networked-agents … #ai #google #agent2age ## 8. The protocol stack: - Agents register with a capability manifest - Handshake est - URL: https://x.com/kkotkkio/status/2045385108782514256 - Author: Working-Ref @kkotkkio - Likes: 0 - Detail capture: not_attempted_detail_limit > The protocol stack: - Agents register with a capability manifest - Handshake establishes trust and message format - Async message exchange handles task delegation Moltbook and Google A2A are building on similar primitives.