June 2026
New Features
Section titled “New Features”MCP Server for AI Assistants
Section titled “MCP Server for AI Assistants”You can now connect Claude Code, Claude Desktop, or any Model Context Protocol client directly to your tenant through the new MCP server. Once connected, an AI assistant can read your risks, incidents, threat profile, and compliance procedures, and — with write access — score risks against the RAMP, triage incidents against the CIRP, assign threat objectives, tag items, and leave comments.
Connecting is a one-time browser sign-in and consent: the assistant inherits your own permissions and never exceeds them, and you can review or revoke any connected application from Settings > Team > Authorized Applications. See the MCP Server guide to get started.
Notable Feature Details
Section titled “Notable Feature Details”-
The MCP server supports parent/child risk and incident hierarchies: an assistant can list the children of a record, create new records nested under a parent, and link or unlink records from a hierarchy — subject to the same single-level nesting constraint that applies in the UI.
-
Freetext search through the MCP uses AND semantics across all terms, so a query like
"local administrator rights"matches records where those words appear anywhere in the title, description, or comments, even if they are not adjacent or fall in different fields. -
The consent screen shows exactly what permissions are being granted, intersected with your current roles — approving a write request as a read-only user is silently narrowed to what your roles actually allow, so the token always reflects your real access. Every connected AI client appears in Settings > Team > Authorized Applications and can be individually revoked at any time. To verify a connection’s identity and permissions, ask the assistant directly — it can report back which organization and user it is acting as, and what it is and isn’t permitted to do in that session.
ServiceNow Ticket Integration
Section titled “ServiceNow Ticket Integration”The platform now connects to ServiceNow to create tickets directly from risk and incident records. Configure the integration in Settings > Integrations using your ServiceNow instance URL and service account credentials. Tickets are created on demand from within a risk or incident record and link back to the originating record for lifecycle tracking. See the ServiceNow integration guide to get started.
Linear Ticket Integration
Section titled “Linear Ticket Integration”The platform now connects to Linear to create issues directly from risk and incident records. Configure the integration in Settings > Integrations using your Linear API key and team identifier. Once configured, Linear appears alongside Jira and ServiceNow in the Tickets section of any risk or incident record — select it to create and link an issue without leaving the platform. See the Linear integration guide to get started.
Parent/Child Relationships for Risks and Incidents
Section titled “Parent/Child Relationships for Risks and Incidents”Risks and incidents now support parent/child groupings, allowing you to nest related records under a single parent. The register displays children collapsed under their parent by default, keeping high-level overviews readable without hiding detail. A single level of nesting is supported — a child record cannot itself be a parent. See Linked Risks and Linked Incidents for details.
Quick Filters on the Risk and Incident Registers
Section titled “Quick Filters on the Risk and Incident Registers”Quick Filters are one-click filters available to all users directly in the filter bar. The platform has built-in filters for the below:
- Overdue — surfaces records past their expected resolution date. Available only for the risk register.
- Assigned to Me — narrows the register to records assigned to the current user. Available on both the risk and incident registers.
- Upgraded — Displays risks whose urgency has increased compared to their initial assessment recorded via the Initially Reported Urgency field. Logic: Current Urgency > Initially Reported Urgency (IRU). This filter is a risk register item only.
- Downgraded — Displays risks whose urgency has decreased compared to their initial assessment recorded via the Initially Reported Urgency field. Logic: Current Urgency < Initially Reported Urgency (IRU). This filter is a risk register item only.
- Show Child Risks / Show Child Incidents - Displays risks / incidents that have been associated to a Parent risk / Parent incident. Prior to searching for a child record, ensure this filter is enabled.
Incremental Integration Sync
Section titled “Incremental Integration Sync”Integration syncs are now incremental — each run processes only records that have changed since the last successful sync, rather than re-fetching the full dataset. This reduces sync duration across all integrations. Sync history for each run is visible in the integration’s job history panel in Settings > Integrations. See the Integrations guide for more.
Pagination Footer on Registers
Section titled “Pagination Footer on Registers”The risk and incident registers now display a pagination footer showing the current page position, total record count, and page navigation controls.
Keyboard Shortcuts for Text Fields
Section titled “Keyboard Shortcuts for Text Fields”Text entry in modals and register cells now follows a consistent set of keyboard shortcuts:
| Key | Text fields (modal + register) | Comment fields |
|---|---|---|
| Enter | Commit and close | New line |
| Shift/Alt + Enter | New line | New line |
| Cmd/Ctrl + Enter | Save and advance to the next record | Submit comment |
| Esc | Abandon changes and close | Cancel |
The Save and advance shortcut is designed for triage workflows — press Cmd/Ctrl+Enter to save a field and move straight to the same field on the next risk or incident, without touching the mouse.
Improvements
Section titled “Improvements”GitHub Integration: Sync and AI Scoring Toggles
Section titled “GitHub Integration: Sync and AI Scoring Toggles”The GitHub Dependabot integration now exposes independent toggles to enable or disable sync and AI scoring separately. This lets you pause automated scoring without stopping the sync, or import vulnerabilities without scoring them until you’re ready.
User Assignment Search
Section titled “User Assignment Search”User picker dropdowns — including risk and incident assignees and member filters — now use server-side search and show only members with the appropriate access permissions. This improves performance in large organizations and ensures assignment pickers only surface eligible members.
Side Navigation Improvements
Section titled “Side Navigation Improvements”Several refinements to the side navigation panel:
- The profile section now shows only your highest-privilege role rather than the full list of assigned roles, reducing clutter for members who hold multiple roles.
- Settings have moved from the tenant name area into the profile section of the side navigation, consolidating account and organization controls in one place.
- The navigation now includes a direct link to docs.adversarial.com, so documentation is always one click away.
- The icon to collapse the side navigation has moved into the navigation area for improved visibility.
Leave Organization
Section titled “Leave Organization”Members can now leave an organization from their profile page. If you are the only active session holder at departure, the platform redirects you to a safe landing page rather than leaving the session in a broken state.