Activity Log

Activity Log is the audit system for recommendation actions, execution outcomes, and reversions.

Expected outcome: You can trace recommendation lifecycle events end-to-end.

Quick Start

1. Open Activity Log.
2. Set date range.
3. Apply optional filters (status/type/actor/campaign/search).
4. Review summary cards.
5. Open row details or run undo where available.

Expected outcome: You can isolate and act on specific events quickly.

Detailed Workflows

Workflow: Root-Cause Analysis on Recent Incident

1. Filter by affected campaign and time window.
2. Narrow by actor and status.
3. Open event details and compare before/after values.
4. Determine if revert is needed.
5. Execute revert or document follow-up.

Expected outcome: Incident analysis produces a concrete corrective path.

Workflow: Undo Applied Recommendation

1. Locate applied event.
2. Trigger undo.
3. Confirm action.
4. Refresh event list and verify status.

Expected outcome: Reversion is executed and recorded in audit trail.

Workflow: Share Filtered Audit Context

1. Apply precise filters.
2. Copy URL.
3. Share with team/support.

Expected outcome: Other users can open the same filtered audit context without manual recreation.

Workflow: Weekly Quality Review

1. Filter last 7 or 30 days.
2. Compare applied/failed/undone distributions.
3. Investigate high-failure types.
4. Feed findings into Auto-Apply tuning.

Expected outcome: Audit data informs proactive reliability improvements.

Common Errors

Error: No events shown

1. Expand date range.
2. Clear filters.
3. Confirm data context (account/marketplace).

Expected outcome: Empty view is validated as real inactivity or corrected filter scope.

Error: Undo unavailable

1. Check action status and eligibility.
2. Confirm event type supports revert.
3. Use manual corrective action when undo is not supported.

Expected outcome: Operator knows whether recovery is automated or manual.

Error: Event details appear incomplete

1. Refresh list.
2. Re-open event.
3. Capture request/event identifiers for escalation if still incomplete.

Expected outcome: Data gap is confirmed and escalated with actionable metadata.

FAQ

Can I separate manual vs automated actions?

Yes. Use actor filter to isolate auto-apply and manual activity.

Is this only for support teams?

No. It is an operator tool for daily execution assurance.

What is the best default range for routine checks?

Use 7 days for daily/weekly operational checks; expand for trend audits.

Last Updated

• Last updated: 2026-02-21
• Version assumptions: AuditPage filter model, summary cards, table actions, and undo flow