Fetch missed events with GET /v1/events

Software on a private network, behind NAT (for example an Odoo installed in an office), cannot receive webhooks. It polls events with GET /v1/events instead. This is also how you recover events missed while a webhook was failing.

How to poll

1. Call with after=0 the first time.
2. Handle the returned events in order, then store next.
3. Next time, call with after=<next>. If the page was full, call again at once; otherwise wait for the next run.

curl "https://api.hubz.vn/v1/events?after=0&limit=100" \
  -H "Authorization: Bearer hbz_live_xxx"

{
  "events": [
    {"id": "1042", "type": "message.created", "version": 1, "createdAt": "2026-11-06T09:15:02Z", "data": {"…": "…"}},
    {"id": "1043", "type": "conversation.updated", "version": 1, "createdAt": "2026-11-06T09:15:40Z", "data": {"…": "…"}}
  ],
  "next": "1043"
}

Parameters

The key needs the events scope. Events are kept for 7 days. Event ids increase: handle each id once.

Python example

import time

import requests

API = "https://api.hubz.vn"
HEADERS = {"Authorization": "Bearer hbz_live_xxx"}


def poll(after: str = "0"):
    """Fetch every new event once; store `after` between runs (events are kept 7 days)."""
    while True:
        res = requests.get(f"{API}/v1/events", headers=HEADERS, params={"after": after, "limit": 500}, timeout=30)
        if res.status_code == 429:
            time.sleep(int(res.headers.get("Retry-After", "60")))
            continue
        res.raise_for_status()
        page = res.json()
        for event in page["events"]:
            handle(event)  # your code: idempotent, keyed by event["id"]
        after = page["next"]
        if len(page["events"]) < 500:
            return after