agentleFS
Sign inStart free
MicrosoftKnown publishermicrosoft/Building-AI-Agents-From-Zero-To-Production/.github/skills/multi-agent-a2a/SKILL.md

multi-agent-a2a

Skill194 starsChanged 4 months agoMITView on GitHub

Help a learner compose agents as networked services using the open Agent-to-Agent (A2A) protocol — expose an agent as an A2A server and consume a remote agent as a peer. Use for Lesson 7 tasks: multi-agent orchestration and A2A client/server wiring.

  • Reads credentials

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Agents as services: instead of one process holding every agent, each agent can be exposed over A2A and called remotely — enabling reuse, …

    line 20
  2. Very specific:

    Start the server, then run the client.

    line 32
  3. Distinctive:

    Keep the preview caveat visible when advising on production use.

    line 34

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Multi-agent & A2A

Guidance for composing agents over the Agent-to-Agent (A2A) protocol (Lesson 7).

Guardrails (always apply)

  • Terminology: A2A (Agent-to-Agent) — an open protocol for agents to call each other as peers.
  • Uses preview packages: agent-framework-a2a, a2a-sdk, and uvicorn (pinned in requirements.txt with a preview note). Present these as preview, not GA.
  • Config via .env: A2A_HOST, A2A_PORT, A2A_SERVER_URL. Models gpt-5.1 / gpt-5-codex.

Concepts a learner should grasp

  1. Agents as services: instead of one process holding every agent, each agent can be exposed over A2A and called remotely — enabling reuse, independent scaling and cross-team composition.
  2. Server vs client: the server publishes an agent (with an agent card) over A2A; the client discovers and calls it as a peer using A2AAgent.
  3. Orchestration: a coordinating agent can delegate sub-tasks to specialised remote agents.

Worked example in this repo (validated live)

  • lesson-7-multi-agent-a2a/a2a_server.py — exposes a Foundry agent over A2A (default port 9000).
  • lesson-7-multi-agent-a2a/a2a_client.py — consumes the remote agent as a peer and prints its response. This pair was validated end-to-end against a live Foundry agent.

How to help

  • Start the server, then run the client; confirm A2A_SERVER_URL matches the server's host/port.
  • Adapt the server's underlying agent (instructions/tools) rather than rewriting the A2A plumbing.
  • Keep the preview caveat visible when advising on production use.

Validate

  • python -m py_compile on both server and client.
  • Client should reach the running server and return a real model response.

References

  • lesson-7-multi-agent-a2a/README.md

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

Reviews can't be read right now.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free