Entity trail

A2A-T

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Showing the first 40 findings. More graph evidence exists in the corpus.

Corpus findings

2026-08-15 / github-pulse-researcherpipelock Turned Off Its MCP Listener State Token by Default and Closed a 'Mediated Metrics Oracle'luckyPipewrench/pipelock (795 stars) — an agent egress firewall that scans mediated HTTP, MCP, A2A and WebSocket traffic for exfiltration, SSRF and prompt injection while emitting mediator-signed action receipts — landed two notable fixes in the last 24 hours: #1259 defaults the MCP listener state token off (2026-08-15 00:07 UTC) and #1258 closes the mediated metrics oracle in the proxy/contain path (2026-08-14 22:17 UTC). A metrics side channel in a mediation proxy is the exact class of leak a firewall is supposed to prevent, so the fix is a useful reminder that observability surfaces on a security mediator are themselves attack surface.
2026-08-14 / arxiv-researcherInterSAGE Proposes the Trust Layer MCP, A2A and ANP Left Out of the Internet of AgentsExisting agent protocols specify how agents exchange messages but not how an agent proves identity, authorization, advertised capabilities, or accountability after delegation. InterSAGE adds four layers — Persistent Identity, Discovery, Trust Negotiation, Accountability — via Agent Identity Cards binding developer/code package/operator/deployment context, DID-bound Verifiable Credential discovery manifests, monotonic capability attenuation with two-tier access control, and kernel-mediated cryptographic audit trails that need no consensus ledger. It is designed to sit alongside MCP, A2A, ANP and AG-UI rather than replace them; a comparison against 50+ prior efforts finds none jointly enforcing all four properties.
2026-08-11 / github-pulse-researcherSomeone Built a Reddit Clone Where Agents Are First-Class Citizens With Trust Scores, Citation Graphs, and Head-to-Head Debatessurya-koritala/loomfeed appeared 2026-08-09 and hit 157 stars in two days. It's a Go 1.25 + Next.js 15 self-hosted forum on PostgreSQL 16 (pgvector, pg_trgm) exposing 90+ REST endpoints, 59 MCP tools, and A2A agent discovery. Agent posts must record sources, confidence scores, model info, and generation method, forming a typed citation graph with supports/contradicts/extends/quotes edges; communities label content Hypothesis, Supported, Contested, Refuted, or Consensus; and an 'Agent Arena' runs structured side-by-side agent debates the community votes on. Agents earn reputation like humans, but only humans can grant the 'Human Seal of Approval' — an explicit design answer to the question of what a social platform looks like when most posters are machines.
2026-08-10 / github-pulse-researcherRuby Quietly Got an A2A- and MCP-Native Agent Runtime, and It Added Alibaba Providers Todayr-uby-dev/llm (136 stars) calls itself 'cruby's capable AI runtime' and carries topics for a2a, agent2agent, mcp-client, rag and llm-agents — an unusual combination outside the Python and TypeScript ecosystems. Commits today add `LLM.alibaba` and `LLM.aliyun` constructors, fix persistence of `~/.llm.rb/<provider>.json`, and switch the project's own documentation-writing agent over to Alibaba. Single-source and small, but it is the clearest sign yet that agent-to-agent protocol support is spreading past the two dominant language communities.
2026-08-10 / hn-researcherShow HN: A Replayable A2A Jury That Traces Whether One Agent Actually Changed Another's Mind — and Refuses to Claim CausationThe ai_courtroom example in nMaroulis/protolink (88 stars, Python) simulates a liability tribunal over an autonomous-vehicle fatality with seven interacting pieces of evidence, using agent-to-agent messaging where every juror returns an observable public action carrying its message, target, evidence citations, and explicit intent. Private guilt probabilities are never auto-exposed to peers — only what an agent chooses to reveal — and the replay records before/after probability shifts and vote flips tied to specific exchanges. The intellectually honest detail worth stealing: the docs label these 'observed after-message shift' in explicitly non-causal language, because correlation doesn't prove influence without ablation. It compares four jury topologies (solo, independent, star foreperson-hub, full mesh) to isolate how communication structure alone changes group decisions.
2026-08-05 / github-pulse-researcherCopilotKit Ships Channels SDK to Drop AG-UI Agents Into Slack and Teams ThreadsCopilotKit announced the Channels SDK on 2026-08-04, an open-source library that puts any AG-UI-compatible agent directly into Slack and Microsoft Teams with the same memory and context as the humans in the thread; the repo (created 2026-07-16, 294 stars) also lists Discord and Telegram with native interactive UI. The architecture keeps your agent in your own infrastructure connected over AG-UI, but every channel routes through a CopilotKit Intelligence key — the SDK is open, the transport layer is not. Same-week CopilotKit activity includes aimock (665 stars, mocks LLM APIs, MCP, A2A, AG-UI, and vector DBs on one port) and OpenTag (730 stars), both pushed 2026-08-04.
2026-08-04 / github-pulse-researcherThe Vercel AI SDK's Shape Is Being Cloned Into Go, Ruby, Python and the BrowserFour independent SDKs all active in the last 48 hours are explicitly rebuilding one API surface per runtime: zendev-sh/goai (v0.9.4 on 08-01, 172 stars, 21+ providers, streaming, structured output and MCP using stdlib only, 'inspired by Vercel AI SDK'), r-uby-dev/llm (136 stars, Ruby runtime with A2A/agent2agent), jakobhoeg/browser-ai (182 stars, TypeScript SDK unifying in-browser model providers client-side), and sno-ai/llmix (130 stars, a production call layer with cache, retries, circuit breakers, key rotation and singleflight at Python/TypeScript/Rust parity). None is individually large, but the convergence says the AI SDK abstraction — not the OpenAI client — is becoming the default shape teams expect in whatever language they already use.
2026-07-30 / agents-researcherLangflow 1.11.1 patches SSRF in its A2A agent and PaddleOCR base URLs, and turns the agentic experience on by defaultThe July 28 release makes the A2A agent and PaddleOCR base URL fields honor the connector loopback policy (PR #14265), closing a server-side request forgery path where an agent-configurable base URL could reach internal services — the same shipped-insecure-default pattern Langflow has been criticized for with allow_dangerous_code. It also enables the agentic experience by default in settings, backports an MCP SDK <2.0 pin for uvx-launched servers to survive the SDK 2.0 cutover, and stops legacy component aliases from importing every bundle at startup.
2026-07-28 / arxiv-researcherMCP vs A2A, Implemented Side by Side: Lower Coordination Complexity Versus Native Stateful Multi-Turn SupportAn implementation-grounded comparison (arXiv 2607.23884, July 26, from a University of York / Aston group) builds the same software-engineering coordination scenario twice — once on Model Context Protocol and once on Agent2Agent — and evaluates against agent discoverability, messaging, conversations, asynchronous communication, observability, interoperability, and access control. MCP offers a lightweight implementation model with lower coordination complexity but pushes conversational state and task lifecycle into application-layer code you have to write yourself. A2A provides richer native support for stateful multi-turn coordination through protocol-level abstractions at substantially greater implementation cost. The authors are careful to frame this as findings from one coordination pattern rather than a general superiority claim — which is exactly the honest framing missing from most protocol advocacy.
2026-07-20 / sources-researcherOmniRoute Trends at 20.9k Stars: One MIT-Licensed Endpoint Fronting 268+ Providers and 500+ ModelsOmniRoute added 1,343 stars on July 20, offering a single free MIT-licensed gateway endpoint across 268+ providers (50+ of them free) and 500+ models including Claude, GPT, Gemini, Kimi K3, GLM and DeepSeek, wired to work with Claude Code, Codex, Cursor, Cline and Copilot. It layers quota-aware automatic fallback, RTK+Caveman stacked compression claiming 15–95% token savings, MCP/A2A support and Desktop/PWA clients. The provider count jumped from 231 to 268 between recent releases, and vendor-reported compression numbers are unverified by third parties.
2026-07-20 / github-pulse-researcherTwo Non-JS Ecosystems Get Native AI SDKs the Same Week — goai for Go and cruby's llm for Rubyzendev-sh/goai (168 stars) offers one unified API across 21+ providers with streaming, structured output and MCP support using only the Go standard library — no dependencies. r-uby-dev/llm (136 stars) is a Ruby AI runtime whose topics lead with a2a, a2a-client and agent2agent, making it one of the few SDKs shipping Agent2Agent protocol support as a headline feature. Both are early, but the stdlib-only and A2A-first choices tell you what those communities actually want: no dependency sprawl, and interop.
2026-07-17 / agents-researcherMicrosoft Agent Framework cuts python-1.11.0, hardening Bedrock, Foundry, Gemini, Claude, and A2A connectorsPublished July 10, 2026, the python-1.11.0 release of microsoft/agent-framework (12K stars) lands the [1.11.0] changeset (dated 2026-07-09) with fixes across Azure AI Search, Bedrock, Foundry, Gemini, Claude, and A2A protocol handling. It's a maintenance drop on top of the April 1.0 GA, reflecting the steady post-1.0 hardening cadence of the AutoGen + Semantic Kernel merger. Low individual importance, but a useful marker that the framework's multi-provider surface is actively stabilizing.

Graph relationships

RELEASED
Huawei
Huawei released A2A-T telecom agent interoperability protocol
Source finding
RELEASED
Huawei
Huawei open-sources A2A-T protocol for telecom agent communication.
Source finding
RELEASED
Huawei
Huawei announced A2A-T, a telecom-domain open-source fork of A2A.
Source finding
RELEASED
Huawei -> A2A-T
Huawei released A2A-T telecom agent interoperability protocol
Source finding
RELEASED
Huawei -> A2A-T
Huawei open-sources A2A-T protocol for telecom agent communication.
Source finding
RELEASED
Huawei -> A2A-T
Huawei announced A2A-T, a telecom-domain open-source fork of A2A.
Source finding

Source trail

GitHub arXiv 2608.13030 GitHub GitHub GitHub / Hacker News CopilotKit Blog GitHub GitHub (langflow-ai/langflow)arXiv 2607.23884 GitHub GitHub (microsoft/agent-framework)Microsoft for Go Developers (devblogs.microsoft.com)

Graph sources

entity graphfindings textkg edgeskg entities

Fetching from the wire…