a2a-topics skillA
a2a-topics is agent-read markdown (skill) from gke-labs/kube-agents: Read and write the agent blackboard - the durable, named topics on the A2A bus where agents publish what they currently know (upgrade readiness, the environment blueprint, dated annotations). Use when asked about fleet or environment state that another agent may already have assessed, before starting a fresh diagnosis, and to record a finding other agents should start from. Do not use for task status (that is stream replay, not a topic), for anything not in `a2a topics list`, or on an install wh.
Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.
What the file says
# A2A topics: the standing state other agents left you A topic is a durable, named place on the bus where one agent publishes what it currently knows, so the next question starts from standing state instead of a cold diagnosis. Tasks are conversations; topics are the blackboard. Two things this skill exists to prevent. The first is you re-deriving an answer another agent already published — a fleet upgrade sweep costs real minutes and real tokens, and the verdict may be sitting on `upgrade-readiness`. The second is you presenting standing state as if you had just established it. Every entry carries who wrote it and when, and you relay both. ## Prerequisites The bus address and identity come from the environment: `NATS_URL` and `A2A_BUS_USER`. The credential itself is a file the operator projects into this container; `a2a` finds it without being told. If `NATS_URL` is unset, this install is not running the A2A bus — say so plainly and answer the question another way. Do not attempt to guess an address. The client is `a2a`, on `PATH` (`/usr/local/bin/a2a`, shipped in the agent image alongside this skill). Because it is on `PATH` rather than at a …
Read the whole file at its exact version.
How to install
mdr add gke-labs/kube-agents/a2a-topics@git:20260918.965b691mdr add gke-labs/kube-agents/a2a-topics@sha256:9a164c18773770a2Pin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.
[](https://markdownregistry.com/a/art_akvvarktdzlauftt)
1 badge views in 30 days
Versions
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20260918.965b691 latest | 2026-09-18 | 965b691 | 5,915 B | A | view · diff |
| git:20260911.dfb1428 | 2026-09-11 | dfb1428 | 5,814 B | A | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (5915 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
gke-labs/kube-agents · 64 stars · license Apache-2.0 · pushed 2026-09-23 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_akvvarktdzlauftt GET https://markdownregistry.com/api/v1/resolve?ref=gke-labs/kube-agents/a2a-topics GET https://markdownregistry.com/api/v1/blob/9a164c18773770a21c524c191d607055b8abbdec7a80e44d5e4b749dd138db17
Your agent does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.