video1.0<iframe src="https://www.loom.com/embed/2781b35242534eeb8764fdc20fa8bca8" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/2781b35242534eeb8764fdc20fa8bca8-90fcfea476629d9f.gif1163.762This Loom explains how Clay account agents handle credit cost, observability, and keeping account fields current. The speaker notes that credit usage varies based on how much context the agent must reason over, and that narrowing prompt fields can reduce context and credits while scoping up increases usage. They emphasize that account agents are not a black box by showing agent run details step by step, including what data they read and actions taken, down to how each agent-managed field was created. Finally, they describe how new signals such as Gong calls or job changes are matched against existing account knowledge in the background to update account fields and optionally sync to the CRM, with controls available in run settings for auto-enrich and recurring schedules.