MemorySync
Debugging / Escalation

Support & Escalation

A good report gets a fast answer. Build the evidence pack below, and leave user content out of it.

Builder

Build the evidence pack

Fill in what you have. The request id is the single most useful field — it identifies your exact call without any user data.

Paste this into your ticket
Operation:     memory.query
Status:        500
Request ID:    <X-Request-ID from the response>
First seen:    <UTC timestamp>
Frequency:     <every call | intermittent | one-off>
Scope:         <project id, and whether it is one end user or many>
Expected:      <what you expected to happen>
Observed:      <what happened instead>
Already tried:  <the checks from the relevant debugging page>
Checklist

Before you send it

  • Remove user content. Identifiers, statuses, and timestamps are enough to investigate.
  • Never paste an API key. If one has been shared anywhere, rotate it first.
  • Say whether it is every call or intermittent — that single detail changes where to look.
  • Include what you already ruled out, so nobody repeats your work.
  • For anything user-visible, say how many end users are affected.
Was this page helpful?