Submitting a Support Ticket

You open support tickets from the support messenger in the Portal. Any member of your organization can file one. An AI assistant you connected can also file one on your behalf.

From the messenger

1. Select Support in the Portal header, on any signed-in page. The launcher in the page corner opens the same messenger.
2. Start a conversation and describe the issue — what you were doing, what you expected, and what happened instead.
3. Send it.

The composer, its fields, and its attachment control belong to Featurebase, so their exact layout can change without a StackJack release. See Ticket attachments for what StackJack can and cannot state about attachment limits, and for what to check before you attach a file.

Priority is not a field you set. The widget does not ask you to choose one, and StackJack records every widget-filed ticket at Medium in its own copy. Say in the description how badly you are blocked; support triages from that. The one path that still carries an explicit priority is the AI-assistant tool described in Support through your AI assistant.

What StackJack does when your ticket arrives

StackJack keeps its own record of the ticket alongside the conversation, so support can see your organization's history and your AI assistant can read the thread. Three things happen:

• StackJack matches the ticket to your organization, which is why filing while signed in matters.
• It records the subject and description, and attaches a diagnostics bundle — for a signed-in submitter only. See the note below.
• It keeps its copy of the conversation up to date as the ticket goes on, so nothing is lost if a notification goes astray.

StackJack sends no confirmation email of its own for a ticket you file in the messenger. Featurebase sends the ticket notifications — see Replies and notifications.

Automated research does not start when you file a ticket. A StackJack operator starts it deliberately, and only they see the result. That is covered on the replies page.

Diagnostics are attached automatically

StackJack attaches a diagnostics bundle to a new ticket so support can start from the actual state of your organization. The bundle can include:

• The organization and ticket submitter's account identifiers and access state.
• Team-member and pending-invite details, including email addresses, identity-provider IDs, roles, status, expiry dates, and assigned tool permissions.
• Subscription plan and usage-limit information, plus up to 100 recent tool-call records from the preceding 48 hours. Records can include connector and tool names, success or failure, timing, and error messages.
• Connector-credential metadata: connector and authentication type, scope, owner, enabled/valid state, validation timestamps, validation errors, and disable reason. Stored credential values are not fields in this report.
• MCP client metadata: client and user IDs, active/authentication/activity timestamps, the linked credential ID, and allowed-tool assignments.
• A separate recent MCP server log bundle when that collection succeeds.

The MCP log collector applies pattern-based masking to common credentials and tokens. Other diagnostic fields can contain original error text, so StackJack does not promise that every sensitive-looking value in every bundle has been removed. Review what you write and what you attach as well; those go through as written.

The bundle is collected once per ticket, when StackJack first records it. A later reply does not mint a new one.

Sign in before you file, or you get no bundle. The bundle contains your organization's own data, so StackJack attaches it only when it can prove who filed the ticket — that is, when the messenger identified you through StackJack sign-in. If you write in anonymously and StackJack can only match you by the email address you typed, it declines to attach the bundle and support starts without it. Automated research is refused on the same ticket for the same reason. Filing from a signed-in Portal page avoids both.

The bundle is stored separately from the conversation and is not shown in it, and it is not returned to an AI assistant that filed the ticket. The people who can retrieve it are signed-in members of your organization, StackJack support staff, and your organization's current managing support partner, if you have one. Support automation can also use diagnostic evidence when researching the ticket; that can involve external processing described in External Data Processing.

Diagnostics collection is best-effort: if any part of it fails, your ticket is still recorded normally. When a ticket is closed, StackJack deletes its diagnostic files and any attachment files it stores for that ticket. See Customer Data & Privacy before requesting organization deletion.

Other ways tickets get created

Two automated paths also open tickets, and both land in the same Featurebase workspace:

• An AI assistant you connected can file tickets on your behalf using StackJack's built-in support tools — for example, after a tool call fails. This works out of the box on any connection. A ticket filed this way does send you a "we received your request" email from StackJack, because it goes through StackJack's own ticket-create path first.
• A StackJack automation can open a ticket when a run hits an error, but only if someone has switched that on for that automation. New automations start with it off, and by default an automation's tickets carry a reduced diagnostics bundle with identity details removed.

See Support through your AI assistant for both paths, including the per-automation settings and the duplicate and volume limits.