Audit Logs is the human-side counterpart to Agent Ledger. Every operator action across the platform: logins, role changes, kill-switch presses, approvals, config edits, exports. Hash-chained for tamper evidence. Searchable across 8 facets. Exports formatted for SOC2, HIPAA, ISO 27001 reviewers.
Logins, MFA challenges, role changes, password changes, API token issuance, kill-switch presses, approvals, config edits, retention changes, exports, support session activity. If a human did it, it is in the log. The page also captures system actions (auto-demotions, kill-switch firings from policy) so the log is the complete operational record.
Audit Logs share the hash chain with Agent Ledger and Decision Bundles. Every entry includes the hash of the previous entry. Editing any past entry breaks the chain at that point and is detectable. Exports include the chain proof. Auditors can independently verify with the same nova-verify CLI used elsewhere.
Eight facets: actor (user or system), action (kill, approve, edit, etc.), tenant, scope (which agent / service / resource), time, source IP, severity, and free-text against the entry body. Stack any combination. Saved searches cover common audit queries: "all kill-switch presses this quarter," "all role changes by sarah."
Pre-formatted exports for SOC2, HIPAA, and ISO 27001 reviews. Each export includes the audit log entries for the time window, the chain proof, and a manifest matching the framework's expected categories. Email-friendly URL with 30-day expiry. The auditor downloads, runs nova-verify, and starts reading.
Subscribe to Nova AI Ops on YouTube for demos, tutorials, and feature deep-dives.
Agent Ledger covers AI; Audit Logs covers humans. Together they tell the full operational story.