Iron A record reviewers can follow
Move from the event into chain metadata, linked entities, the run, and the recorded evaluation trace.
Audit, Replay & RecoverySee what happened.
See what happened.
Compare what could change.
Search the activity record, inspect the policy and evidence behind a decision, and compare retained historical facts with a controlled simulation. Review cited governance narratives before they are published or exported.
How it works
Recovery starts with understanding the recorded decision.
Open the audit event, follow it to the agent run, and use retained responses or an eligible non-production connector shadow to compare a proposed policy or configuration change. The comparison shows what the historical record says and what the simulation returned.
Product Walkthrough
Trace an event into a replay comparison and reviewed narrative.
See how an operator traces an event, compares a controlled alternative with the historical record, and reviews an explanation tied to evidence.
Historical fact beside simulation
Compare retained evidence with a controlled candidate without presenting the simulation as a rerun of live work.
Narratives tied to evidence
Review claims, fact IDs, evidence IDs, caveats, validation, and approval state before publication.
Why it matters
What this changes once the work gets real.
Unavailable evidence stays visible
A comparison can report unavailable retained evidence instead of filling the gap with an assumed result.
Candidate connectors stay non-production
Eligible connector shadows are attested for read-only or dry-run comparison rather than live customer actions.
Recovery does not rewrite history
Operator recovery and compensation steps remain separate from the original evidence and do not promise that prior effects can be undone.
Make the next decision from evidence, not reconstruction by memory.
Trace the record, compare a controlled candidate, and keep reviewed explanations connected to their sources.