Audit Logs

Every significant action in QuantAssure is logged. When an auditor asks who did what and when — including what the AI decided — this is where you answer.

What's Logged

  • All create, update, and delete operations
  • Status changes on findings, actions, and controls
  • User authentication events (login, logout, failed attempts)
  • Risk acceptances: when one is made, when it ends and when it expires. Dismissals are logged as status changes on the finding
  • Report generation
  • System configuration changes. A configuration change records which fields changed, never their values. Connector credentials (tokens, passwords, keys, webhook URLs) are never shown in the audit log: where an older entry recorded one, it appears as [REDACTED]
  • Team management actions (invites, removals)
  • Integration sync events
  • AI triage events
  • Every MCP tool call and resource read made with a support key (support.access), shown as Quant Support with the support case reference. Support access lists these events on their own

AI Decision Auditability

Every AI severity reclassification is logged as a finding.severity_reclassified event — so when AI adjusts a finding from its raw scanner severity, the audit trail records what changed and when. An event is written only when the effective severity changes: it records the previous and new effective severity, with the raw scanner severity kept as context. Re-triaging a finding that keeps the same adjusted severity writes no new event.

These events appear in the audit log with a readable label in the grey action badge on each row (for example Severity reclassified for finding.severity_reclassified). The row's headline is a plain description, such as "AI reclassified severity of finding …". To review every adjustment the AI has made, set the Action filter to Severity reclassified; Severity overridden and Severity override cleared list the changes people made.

Assigning or unassigning a finding's owner writes finding.assigned ("Changed owner of finding …") with the previous and new owner.

When a person overrides an AI severity, the audit log records finding.severity_overridden ("Overrode severity of finding …") with the previous and new severity, the AI's own severity, the SLA change and the reason, and finding.severity_override_cleared ("Reverted severity override on finding …") when it is reverted. If the AI re-analyses an overridden finding and its severity changes, a finding.severity_override_ai_changed event is written; the override stays in force.

Risk acceptances are logged against a risk acceptance entity. Making one is an Approve action ("Risk accepted for finding "…": reason. Expires …") that records the reason, justification and expiry. When it ends, a Status Change entry reads "Risk acceptance ended" with the reason it ended. An acceptance that runs out is logged as risk_acceptance.expired ("Risk acceptance expired"). Choose Risk acceptance in the Entity Type menu to list them, or set the Action menu to Risk acceptance expired to see the ones that ran out. The finding's own status change is logged too.

For a point-in-time evidence export of every AI decision — original vs adjusted severity, reasoning, and confidence — generate the AI Decision Audit Trail report.

Viewing Audit Logs

  • Navigate to Reports & audit → Audit log
  • Filter by:
    • Entity Type: Finding, Policy, Action, System, Compliance Target, User, Organisation, Integration, Risk acceptance, Report, Run, Agent Job or Connection Check
    • Action: Create, Update, Delete, Status change, Assign, Approve, Reject, Sync, Export, Login, Logout or Support access, or one of these specific events: Finding assigned, Severity reclassified, Severity overridden, Severity override cleared, Source removed, Source restored or Risk acceptance expired
    • From and To dates (both days are included)
    • Search by entity name or description
    There is no filter for who made the change. Their name is on each row.
  • Each row shows a description of the event, who made it (marked Agent for an automated agent or Support for Quant Support), the action, the entity, the IP address when one was recorded, and the time. The list is 50 events a page. Open a row to see the entity (with a link to it where one exists), who made the change, the full description and the IP address

Exporting for Compliance

  • Click "Export CSV" to download the filtered audit trail
  • CSV exports are useful for IRAP evidence and internal audits
  • Include date range filters to scope exports to audit periods
  • The file holds every event matching the current filters, not just the page on screen. An export is limited to 50,000 events; if more match, narrow the date range
Tip: Export your audit trail as part of your IRAP evidence pack. Assessors value a clear record of security-related activities.