A2A (Agent2Agent) protocol
Tied for most recommended of 7 agent communication protocols
Open standard letting separate AI agents talk to each other; praised for point-to-point links. Critics call it overcomplicated and say it skips the harder problem.
Open standard letting separate AI agents talk to each other; praised for point-to-point links. Critics call it overcomplicated and say it skips the harder problem.
35 people gave an opinion
- recommend it
- 14
- criticize it
- 21
- approval share
- 40%
Recommended in 5 of 13 months
- November 2025: recommended
- December 2025: not recommended
- January 2026: not recommended
- February 2026: not recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: not recommended
- June 2026: recommended
- July 2026: not recommended
- August 2026: not recommended
- September 2026: not recommended
- October 2026: recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 14 people recommend itNeeds 10 or more, met.
- 0.6 recommenders for every criticNeeds 2 to 1 or better, not met.
- Recommended in 5 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Fits Google ecosystem well
“The only down side I see to the google approach is TLS from a verified CA, rotation of certs is always a pain.”
Comment from 1 Nov 2025 in “How are you handling identities for AI agents?” “If you're big in the google ecosystem it makes sense”
Comment from 18 Jun 2026 in “Ask HN: Is anyone using the A2A protocol yet?” Good agent-to-agent communication
“The A2A protocol is a good (not perfect) solution for agent-to-agent communication”
Comment from 19 Jun 2026 in “Ask HN: Is anyone using the A2A protocol yet?” “works well for point-to-point connections, but it skips the harder problem”
Comment from 7 Apr 2026 in “Show HN: M2a – Chat with remote agents from your terminal” Expected to matter going forward
“going forward stuff like MCP and A2A are way more relevant”
Comment from 11 Jun 2026 in “The Road to the WASM Component Model 1.0” “enterprises are, its the future for them”
Comment from 18 Jun 2026 in “Ask HN: Is anyone using the A2A protocol yet?”
What people warn about
Each mark is one comment
Overcomplicated for most use cases
“we get shitty things like A2A or MCP or whatever the next "shiny" overengineered thing is”
Comment from 18 Jun 2026 in “Ask HN: Is anyone using the A2A protocol yet?” “Seems like over engineering just build an API in front of your agent”
Comment from 18 Jun 2026 in “Ask HN: Is anyone using the A2A protocol yet?” “A2A didn't fit the use case of "trusted agent, and was bloated”
Comment from 18 Jun 2026 in “Ask HN: Is anyone using the A2A protocol yet?” - Comment from 18 Jun 2026 in “Ask HN: Is anyone using the A2A protocol yet?”
- Comment from 20 Aug 2026 in “A2A: Adversary to Adversary?”
Little real-world use or momentum
“it doesn't feel like it will get the momentum it needs to take off”
Comment from 18 Jun 2026 in “Ask HN: Is anyone using the A2A protocol yet?” “nope, not a single use case i witnessed used a2a in the final product”
Comment from 18 Jun 2026 in “Ask HN: Is anyone using the A2A protocol yet?” “don't see the point of A2A except some really specific cases”
Comment from 4 Sep 2026 in “Ask HN: Who is using MCP in production?” Misses the core problem
“works well for point-to-point connections, but it skips the harder problem”
Comment from 7 Apr 2026 in “Show HN: M2a – Chat with remote agents from your terminal” “the A2A protocol was not solving the correct problem”
Comment from 18 Jun 2026 in “Ask HN: Is anyone using the A2A protocol yet?”
Other agent communication protocols people recommend
The four most recommended other agent communication protocols, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| A2A (Agent2Agent) protocolThis page | 14 | 21 | 40% | 5 | — | — |
| Murmur by InstaVM | 14 | 0 | 100% | 6 | — | Approved |
| GitAgent | 11 | 4 | 73% | 1 | — | — |
| MCP Apps | 10 | 3 | 77% | 5 | — | Approved |
| A2UI | 10 | 8 | 56% | 4 | — | — |
What people said
32 of the 36 comments that mention it, newest first. Each links to the original comment.
“I tested the CLI, it is amazing.”
“don't see the point of A2A except some really specific cases”
“it seems too complicated for ape adoption”
“another seemingly pointless standard with an even worse name”
“Neither x402 Bazaar nor Google's A2A verifies supplier output.”
“A2A will be just more of the same”
“The A2A protocol is a good (not perfect) solution for agent-to-agent communication”
“If you're big in the google ecosystem it makes sense”
“It just felt over complicated for our use case.”
“the A2A protocol was not solving the correct problem”
“we get shitty things like A2A or MCP or whatever the next "shiny" overengineered thing is”
“for the most part, it's back to Agent <> MCP now”
“nope, not a single use case i witnessed used a2a in the final product”
“A2A didn't fit the use case of "trusted agent, and was bloated”
“it doesn't feel like it will get the momentum it needs to take off”
“Seems like over engineering just build an API in front of your agent”
“It's quite challenging to manage.”
“Allows you or teams to develop agents independently and have them interact as and when they need to.”
“was skeptical whether they (including A2A) added much value”
“we realized we didn’t need an agent on the other side so now we have this weird hybrid pattern”
“designed on paper by someone with no understanding of prompt caching”
“enterprises are, its the future for them”
“going forward stuff like MCP and A2A are way more relevant”
“Google's A2A and AP2 are missing a critical protocol”
“works well for point-to-point connections, but it skips the harder problem”
“No framework we could find — AutoGen, CrewAI, LangGraph, OpenAI Agents SDK, Google A2A — answers a basic question”
“MCP and A2A are mostly antiquated”
“platforms already irrelevant like OpenAI Agent SDK, Google A2A”
“Having A2A is much more efficient and less error prone.”
“is the best place to start.”
“A2A would be better than this, then we can get the UI out of the way”
“The only down side I see to the google approach is TLS from a verified CA, rotation of certs is always a pain.”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.