{"slug": "multi-agentic-system", "title": "multi-agentic system", "summary": "A developer built a distributed multi-agent system using Google's open-source Agent Development Kit and the Agent-to-Agent (A2A) protocol, in which a researcher, judge, and content_builder run as independent HTTP services that discover each other via A2A agent cards. The pipeline researches a topic, has the research critiqued and looped back until an editor approves it, then writes a course module, with progress streamed to a browser front end over Server-Sent Events. A single command starts the three A2A servers, runs the pipeline, and shuts them down on exit.", "body_md": "*This post is my submission for [DEV Education Track: Build Multi-Agent Systems with ADK](https://dev.to/deved/build-multi-agent-systems).*\n\nA distributed multi-agent system that researches a topic, has the research\n\ncritiqued, loops until an editor approves it, then writes a course module.\n\nThe Agent Development Kit and the Agent-to-Agent protocol are both open source,\n\nso the multi-agent design itself is unchanged — three agents still run as\n\nindependent HTTP services that discover each other through A2A agent cards.\n\nThis starts the three A2A servers, waits for their agent cards, runs the\n\npipeline, writes results to `output/`, and shuts the servers down.\n\nThe codelab ships a browser front end; this is the local equivalent. One command\n\nbrings up the whole system — the web process starts the three A2A servers on\n\nlaunch and stops them on exit, so there is nothing else to run:\n\n```\n.venv\\Scripts\\python.exe scripts\\run_web.py        # then open http://localhost:8080\n```\n\nType a topic and press **Build the course**. The page shows the three agents as a\n\nlive pipeline diagram: each box is a separate process, and it lights up as its\n\nA2A reply arrives — `researcher` researching → done, `judge` reviewing → approved\n\n(or looping the researcher back for another pass), `content_builder` writing →\n\ndone. An event log timestamps each step, and the finished module renders as\n\nformatted markdown in the \"paper\" pane on the right.\n\nProgress is pushed to the browser over Server-Sent Events\n\n(`GET /api/runs/{id}/events`); a 5-second poll for the finished document is the\n\nfallback if the stream hiccups. The same backend the CLI uses drives it, so a run\n\nhere is identical to `run_all.py` — it just does not write to `output/`.\n\nOther endpoints: `GET /api/health` (model + which agent cards are up) and\n\n`POST /api/runs` (start a run, returns `{\"run_id\": ...}`).", "url": "https://wpnews.pro/news/multi-agentic-system", "canonical_source": "https://dev.to/tejas_rahane/multi-agentic-system-eei", "published_at": "2026-10-11 08:09:54+00:00", "updated_at": "2026-10-11 08:21:36.855581+00:00", "lang": "en", "topics": ["ai-agents", "agent-protocols", "developer-tools", "large-language-models", "ai-tools"], "entities": ["Agent Development Kit", "Agent-to-Agent protocol", "DEV Education Track"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/multi-agentic-system", "markdown": "https://wpnews.pro/news/multi-agentic-system.md", "text": "https://wpnews.pro/news/multi-agentic-system.txt", "jsonld": "https://wpnews.pro/news/multi-agentic-system.jsonld"}}