Our org records outbound calls via automatic recording on a call queue, so
every recording is owned by the queue, not a user. GET /v2/phone/users/me/recordings
returns "owner": { "type": "callQueue" } for all of them, with no user_id
field. The authorizing user is on that queue’s recording access member list.
Q1. Does phone.recording_completed_for_access_member fire in this case?
Or does it only fire for recordings owned by a user who granted access to
others? If it doesn’t fire here, what is the correct event / scope / app type
for an app that needs queue-owned recordings its user can access?
Q2. If it does fire, is payload.object.recordings[].user_id present?
Our integration resolves the owner from that field, and it is absent in the REST
response. Knowing this before we build would save a redesign.
Already verified
- Non-Phone events arrive; Phone events never have. Same app, same endpoint.
Auser.updatedwas delivered and answered HTTP 200, arriving 193 ms
after the payload timestamp. So the endpoint works. (We did have a WAF problem
on our side — found and fixed, and proven fixed by the above.) - All 68 Phone events are subscribed and still nothing arrives, including
recording-independent ones likephone.caller_ringing. Re-authorizing the app
changed nothing. - The calls and recordings exist — both appear in
call_logsand
recordings, with transcript URLs. - Scopes granted:
user:read:user,phone:read:list_call_logs,
phone:read:list_recordings. If Phone webhooks need more, that would explain
it, but I couldn’t find that documented.
Note: the Marketplace webhook logs page shows nothing for this app — not even
the user.updated I can prove arrived — so it can’t be used as evidence either
way. (API Call Logs works fine.)
I can’t test the alternative: outbound calling from the personal extension is
disabled by org policy, and queue-level automatic recording is an org-wide
security requirement. So every recording this account can produce is
queue-owned — meaning if these events exclude queue-owned recordings, that’s a
hard blocker for us, not a test artifact.
Thanks in advance.