Complete operational visibility for your entire org — live process tracking, a full compliance audit trail, and a dedicated AI agent trail that shows every proposal, approval, and execution.
Three lenses on one platform: live process tracking for running operations, a searchable audit trail for everything that ever happened, and a dedicated AI agent log that keeps every autonomous action accountable.
Eight WebSocket channels stream real-time step progress for every long-running operation — provisioning, backups, restores, deployments, SSL, Docker builds, and more. Cancel, retry, or clean up right from the feed.
Every state-changing action in the platform is logged — who did it, from which IP, with which browser, at exactly what time. 65+ action types span infrastructure, security, deployments, backups, billing, and team management.
Every AI proposal, user decision, and execution result is logged separately. You can see exactly what the agent proposed, whether you approved or denied it, and whether execution succeeded — with full metadata and tool parameters.
Long-running operations don't disappear into a spinner. Each one streams live step-by-step progress directly to the Activity Center via WebSocket — so you always know exactly what CloudAIPilot is doing to your infrastructure right now.
Server provisioning, backup/restore, site setup, SSL, Git deploy, Docker app deploy, cloud upload/download, and PQC activation — all stream live with 5-second reconnect on disconnect and fallback polling.
When a step fails, expand it in the process panel to read the full error. Retry failed provisioning jobs or clean up stuck processes without leaving the Activity Center.
Paginated log of all completed operations — backups, restores, site setups, SSL provisions, deployments, migrations. Filter by type, status, server, cloud account, or date range.
The AI Activity tab filters the audit log to show only agent.* actions — forming a dedicated trail of every proposal the AI made, every decision you took, and every execution that followed.
From proposal creation through user decision to execution result — each stage is a separate audit entry. You always have the full chain: what was proposed, what you said, what happened.
Every AI action is tagged: safe, caution, or destructive. Destructive actions that required explicit confirmation text show that confirmation in the log entry.
Every state-changing action in CloudAIPilot writes an immutable audit entry — user ID, IP address, browser user agent, resource type, resource ID, and structured metadata. Search by action name, filter by user, resource type, or date range.
Search by action name substring, or stack filters: action type, actor user, resource type, and date range are all combinable. Filter dropdowns auto-populate with values from your org's actual log.
Every audit row expands to reveal the full structured metadata JSON — backup type, server hostname, error codes, before/after values for updates, tool parameters for AI actions. No guessing what actually happened.
Audit logging is fire-and-forget — it never blocks or slows down the operation being logged. Logs are retained indefinitely. Access is gated by a dedicated permission so only the right people can read your audit trail.
Audit entries are never deleted by the platform. Billing and financial events carry a 7-year compliance window. Personal account security events are accessible for the last 90 days.
Only team members with the org:audit permission can access the audit log. Admins and billing owners get it by default. Custom roles can be scoped precisely.
Audit logging never slows down or blocks the operation it's recording. Failures are silently absorbed — a logging hiccup never takes down a backup job or a deployment pipeline.
Live process feeds, a full audit trail, and a dedicated AI agent log — all in one Activity Center your whole team can trust.
No credit card required · Cancel any time