video1.0<iframe src="https://www.loom.com/embed/d54035b03f66466ea2996218c1a1f646" frameborder="0" width="1280" height="960" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>9601280Loomhttps://www.loom.com9601280https://cdn.loom.com/sessions/thumbnails/d54035b03f66466ea2996218c1a1f646-c0c63b2c92f20d1c.gif585.003This Loom provides a walkthrough of a job-search automation system built as a team of specialized AI agents. The presenter explains the architecture where Slack messages are routed to agents that integrate with tools, write to a shared SQLite memory layer, and power a dashboard with a “morning brief” that refreshes twice daily. They detail scoring across 5 dimensions using weighted rubric scoring with Thompson sampling weights, plus a feedback loop that adjusts future scoring and triggers actions when the composite score is over 70. The build was done in Cloud Code with over 10,000 lines of TypeScript and 52 passing tests, using pre-pushed hooks and extensive testing across API routes, Slack bots, and OAuth flows.