Decision Traces

Decision Traces are immutable records of what happened and why. Every agent decision creates a trace that captures the complete context and outcome.

Overview

Decision Traces are the foundation of TraceMem's auditability. They provide a complete, queryable record of every decision made by your AI agents.

What's Captured

Every decision trace includes:

• The decision envelope (intent, automation mode, actor, lifecycle state)
• All events (reads, writes, policy evaluations, approvals)
• Decision-time snapshots (policy hashes, schema hashes, data product hashes)
• Final outcomes and what changed

Immutability

Traces are append-only and immutable, ensuring that "why this was allowed" remains answerable even years later, even if policies or data products change.

Querying Traces

[Content to be filled - how to query traces, what you can search for]