StackJack 0.1.0.5206+5be0236

This release restores a previously withdrawn Acronis tool, fixes connector problems in GravityZone and Alloy Navigator, gives agents more accurate tool descriptions for Teramind, SentinelOne and Yeastar, and extends how long StackJack keeps trying to bring an auto-disabled connector back online.

New features

• Acronis: acronis_delete_psa_sales_items is available again. It now requires a sales item id, and a call without one is refused before any request is sent to Acronis.

Improvements

• Connector auto-recovery: When a connector is automatically disabled after repeated authentication refusals, StackJack now retries reconnecting every 2 hours for up to 8 attempts (roughly 14 hours), instead of hourly for 3 attempts. Vendor outages lasting most of a day can now recover on their own without manual re-enabling.
• SentinelOne: s1_has_policy is now classified as a Free, read-only lookup. It was previously listed as a Pro, destructive tool with guidance that did not match what it actually does (it only answers whether matching agents have an upgrade policy).
• Teramind: The bodyJson parameter on 16 update/create tools now lists every request-body field the vendor documents, not just the first 12, and duplicates have been removed. The short tool summary still shows up to 12 fields but now says how many more exist. Agents previously composed incomplete request bodies and had them rejected because fields beyond the twelfth were never shown.
• Yeastar: yeastar_update_conference now states clearly that moderator_list replaces the entire moderator list, and explains how to read the current list with yeastar_get_conference and re-send existing moderator IDs alongside new ones. yeastar_create_conference describes the field the same way.

Bug fixes

• GravityZone: Creating a response action that targets an incident by integration identifiers (company and tenant identifiers) now works. The identifiers were previously sent to Bitdefender in the wrong format, so that path of the tool always failed.
• Alloy Navigator: Saving or re-enabling a connector no longer fails when the application account has no personal user record. The credential check now confirms the credentials work without relying on that record, so connectors that previously could not be validated — and could end up disabled with no way to recover — now validate normally.

UI & UX changes

• GravityZone: The connector setup panel and setup guide no longer claim that Test Connection reports which API groups the key holds. The test confirms the key works; to see the granted API groups, use the read-only gravityzone_get_api_key_details tool.
• Alloy Navigator: The setup instructions have been updated to describe what the credential check actually verifies.

Known issues

• Acronis: The request shape used by acronis_delete_psa_sales_items follows the pattern of Acronis's sibling delete endpoints but has not been confirmed against a live tenant. Test it with a throwaway sales item before relying on it.
• Connector auto-recovery: Because recovery attempts are now spaced 2 hours apart, a connector that recovers quickly after being auto-disabled is noticed at about 2 hours rather than 60–90 minutes. For per-user connections that repeatedly fail and recover within the 7-day window, the number of re-authorization emails a user can receive rises from 3 to 8. Shared connections do not send these emails.