DEV Community

MyZucoins
MyZucoins

Posted on Fully Autonomous

Turn MCP search results into an Atom feed of Zucoin wallet releases with Python

If you build on Zucoin, for example a "Pay with Zucoin" link that prefills a transfer code, you want to know when a new wallet app version ships. A feed reader is still one of the simplest ways to follow that. In this tutorial we write a short Python script that asks the open MyZucoins MCP server for its system updates, keeps the wallet app releases, and writes a standard Atom feed that any feed reader can subscribe to. It uses only the standard library.

MyZucoins is a community to accelerate the adoption of Zucoin and its Splitchain network. Its MCP server is read-only published content: it cannot access wallets, sign or send transfers, or read accounts, and it needs no account or API key.

Disclosure: this is written by the team behind MyZucoins, using an AI agent that drafted the post and ran every code sample against the live server with Python 3 on 12 November 2026.

The plan

The script makes these calls to https://myzucoins.org/mcp:

  1. initialize, then the notifications/initialized notification, as the MCP lifecycle describes.
  2. tools/call with search_articles, filtered to the system-news topic, following next_offset until every page is read.

Each search result already has what a feed entry needs: title, excerpt, published_at and url. So there is no need to call get_article for every release.

The JSON-RPC helper

The Accept header must list both application/json and text/event-stream, as the Streamable HTTP transport requires. Set your own User-Agent too. With Python's default Python-urllib agent, the site's firewall answered 403 in our tests, and wallet-releases/0.1 worked.

"""Build an Atom feed of Zucoin wallet app releases from the MyZucoins MCP server."""
import json
import re
import sys
import urllib.request
import xml.etree.ElementTree as ET
from datetime import datetime, timezone

MCP_URL = "https://myzucoins.org/mcp"
HEADERS = {
    "Content-Type": "application/json",
    "Accept": "application/json, text/event-stream",
    "User-Agent": "wallet-releases/0.1",
}
RELEASE_TITLE = re.compile(r"\bwallet (?:app )?v(\d+)\b", re.IGNORECASE)
ATOM = "http://www.w3.org/2005/Atom"
_next_id = 0


def post(body):
    req = urllib.request.Request(
        MCP_URL, data=json.dumps(body).encode(), headers=HEADERS
    )
    with urllib.request.urlopen(req, timeout=60) as resp:
        raw = resp.read().decode()
        if resp.headers.get_content_type() == "text/event-stream":
            raw = [l[5:] for l in raw.splitlines() if l.startswith("data:")][-1]
        return json.loads(raw) if raw else None


def rpc(method, params=None):
    global _next_id
    _next_id += 1
    body = {"jsonrpc": "2.0", "id": _next_id, "method": method}
    if params is not None:
        body["params"] = params
    message = post(body)
    if "error" in message:
        raise RuntimeError(message["error"]["message"])
    return message["result"]


def call_tool(name, arguments):
    result = rpc("tools/call", {"name": name, "arguments": arguments})
    if result.get("isError"):
        raise RuntimeError(result["content"][0]["text"])
    return result["structuredContent"]
Enter fullscreen mode Exit fullscreen mode

post() handles both a plain JSON answer and a server-sent event stream. The MyZucoins server answered with plain JSON, and it answered the notification with 202 and an empty body, which is why post() returns None for an empty response.

Finding the releases

Topic IDs come from list_topics. On 12 November 2026 it listed 60 articles under system-news, which also covers network tools, website updates and other announcements. The script searches that topic for "wallet", then keeps only titles that contain a version number:

def release_articles():
    """Page through system news about the wallet, keep titles with a version."""
    articles, offset = [], 0
    while offset is not None:
        page = call_tool("search_articles", {
            "topic": "system-news",
            "query": "wallet",
            "limit": 100,
            "offset": offset,
        })
        articles += page["articles"]
        offset = page["next_offset"]
    releases = []
    for article in articles:
        match = RELEASE_TITLE.search(article["title"])
        if match:
            releases.append({**article, "version": int(match.group(1))})
    return sorted(releases, key=lambda a: a["published_at"], reverse=True)


def atom_feed(releases):
    ET.register_namespace("", ATOM)
    feed = ET.Element(f"{{{ATOM}}}feed")

    def add(parent, tag, text=None, **attrs):
        el = ET.SubElement(parent, f"{{{ATOM}}}{tag}", attrs)
        el.text = text
        return el

    add(feed, "id", "tag:wallet-releases,2026:zucoin-wallet-app")
    add(feed, "title", "Zucoin wallet app releases")
    add(feed, "updated", releases[0]["published_at"])
    add(feed, "link", href="https://myzucoins.org/")
    add(add(feed, "author"), "name", "MyZucoins")
    for r in releases:
        entry = add(feed, "entry")
        add(entry, "id", r["url"])
        add(entry, "title", r["title"])
        add(entry, "updated", r["published_at"])
        add(entry, "link", href=r["url"])
        add(entry, "summary", r["excerpt"])
    ET.indent(feed)
    return ET.tostring(feed, encoding="unicode", xml_declaration=True)
Enter fullscreen mode Exit fullscreen mode

Atom needs an id, a title and an updated time for the feed and for each entry. The article URL is a stable ID, and published_at is already an RFC 3339 timestamp, which is the format Atom expects.

Putting it together

def main():
    rpc("initialize", {
        "protocolVersion": "2025-06-18",
        "capabilities": {},
        "clientInfo": {"name": "wallet-releases", "version": "0.1"},
    })
    post({"jsonrpc": "2.0", "method": "notifications/initialized"})

    releases = release_articles()
    newest = releases[0]
    age = datetime.now(timezone.utc) - datetime.fromisoformat(newest["published_at"])
    print(f"{len(releases)} wallet app releases found")
    print(f"Newest: v{newest['version']} on {newest['published_at'][:10]} ({age.days} days ago)")
    for r in releases[:5]:
        print(f"  v{r['version']}  {r['published_at'][:10]}  {r['url']}")

    out = sys.argv[1] if len(sys.argv) > 1 else "zucoin-wallet-releases.xml"
    with open(out, "w", encoding="utf-8") as f:
        f.write(atom_feed(releases))
    print(f"Wrote {out}")


if __name__ == "__main__":
    main()
Enter fullscreen mode Exit fullscreen mode

Save the three blocks in order as wallet_releases.py.

Run it

python3 wallet_releases.py
Enter fullscreen mode Exit fullscreen mode

This is the output we got on 12 November 2026:

31 wallet app releases found
Newest: v201 on 2026-10-01 (42 days ago)
  v201  2026-10-01  https://myzucoins.org/zucoin-wallet-app-v201-released-faster-loading-and-wider-device-support/
  v200  2026-08-02  https://myzucoins.org/zucoin-wallet-app-v200-released-transfer-messages-easier-contacts-and-clearer-history-records/
  v199  2026-06-04  https://myzucoins.org/zucoin-wallet-app-v199-released-expiring-transactions-fuzzy-search-and-more/
  v198  2026-03-01  https://myzucoins.org/zucoin-wallet-app-v198-update-released-reliability-and-mini-apps/
  v197  2025-12-20  https://myzucoins.org/zucoin-wallet-app-v197-and-splitchain-update-released/
Wrote zucoin-wallet-releases.xml
Enter fullscreen mode Exit fullscreen mode

And the start of the feed file:

<?xml version='1.0' encoding='utf-8'?>
<feed xmlns="http://www.w3.org/2005/Atom">
  <id>tag:wallet-releases,2026:zucoin-wallet-app</id>
  <title>Zucoin wallet app releases</title>
  <updated>2026-10-01T07:00:00+00:00</updated>
  <link href="https://myzucoins.org/" />
  <author>
    <name>MyZucoins</name>
  </author>
  <entry>
    <id>https://myzucoins.org/zucoin-wallet-app-v201-released-faster-loading-and-wider-device-support/</id>
    <title>Zucoin wallet app v201 released: Loads nearly 50% faster and wider device support</title>
    <updated>2026-10-01T07:00:00+00:00</updated>
    <link href="https://myzucoins.org/zucoin-wallet-app-v201-released-faster-loading-and-wider-device-support/" />
    <summary>With 48.9% fewer items to load and nearly 1,600 changes, Zucoin wallet app v201 brings quicker startup, refined layouts, and better support across more devices.</summary>
  </entry>
  <entry>
    <id>https://myzucoins.org/zucoin-wallet-app-v200-released-transfer-messages-easier-contacts-and-clearer-history-records/</id>
    <title>Zucoin wallet app v200 released: Transfer messages, easier contacts, clearer history</title>
    <updated>2026-08-02T07:00:00+00:00</updated>
Enter fullscreen mode Exit fullscreen mode

The feed has 31 entries, from v171 on 13 February 2024 to v201 on 1 October 2026, with no gaps in the version numbers. Point a feed reader at the file, or serve it from any static host, and new releases show up after the next run.

Two search details that matter

Our first version searched for "wallet app" and only accepted titles that started with "Zucoin wallet app v". It found 29 releases and silently missed two:

  • v197 was announced as "Zucoin releases mega update, ... simpler wallet v197, ...", so the words "wallet app" never appear next to each other.
  • v192 was announced together with another update, so its title ends with "+ Zucoin wallet app v192!" instead of starting with it.

search_articles matches words in the title, excerpt or topic, so a broader query plus a version check in your own code is more reliable than one exact phrase. Checking the version list for gaps is a cheap test that the search found everything.

Also note what this feed is. Each entry is a dated article, not live network status. The server's own instructions describe older articles as historical, so link to the article, as the feed does, and read the release notes there. The v201 release notes, for example, cover faster loading and wider device support.

To run it on a schedule, add it to cron or a CI job and commit the XML wherever your feed reader can reach it. Setup steps for connecting the same MCP server to Claude, Cursor and other clients are at myzucoins.org/mcp.

Top comments (0)