What is an Agent Activity Log?

Unlike raw debug output meant for engineers, an activity log is written for the person who has to answer "what did the agent actually do?" — a business owner checking whether an overnight automation ran correctly, or a compliance officer reconstructing events after an incident. It typically reads as a timeline: the agent read this file, called this API, made this decision, and produced this result, in plain language rather than stack traces. Activity logs are distinct from an audit trail in emphasis: the audit trail is the tamper-evident, compliance-grade record built for regulators and forensic review, while the activity log is the operational, day-to-day visibility layer a team actually reads. Good agent platforms generate both from the same underlying event stream so they never drift out of sync, and let a user drill from a log entry straight into the underlying tool call or approval decision that produced it.