dream

Cross-repo pattern discovery is done by running parallel agents to mine session logs, git histories, and repository listings to surface system-wide problems, plans, gains, and lost work. It categorizes findings into a markdown report with connections and next actions.

Cross-repository triage identifies systemic risks and forgotten work by mining session histories and Oracle memory. Five parallel agents gather diverse signals from your repositories, classifying findings into categories like PAIN, PLAN, GAIN, and LOST to reveal system-wide patterns.

Generate a categorized retrospective report by running a full cross-repo survey using the /dream command. This launches parallel agents to collect signals across project repositories, Oracle memory, and session traces, outputting an append-only markdown report with insights and next actions.

Yes, multi-agent analysis mines git histories and recent activity across all project repositories to identify abandoned work. It classifies these discoveries as LOST items, connecting them with other findings to help you make informed recovery and prioritization decisions.

Cross-repo pattern discovery uses Oracle memory as one of five parallel agents to gather learned memories and fleet status. While it applies across session histories and git logs, Oracle memory provides essential context for surfacing systemic risks and the emotional state of the system.

Surface systemic risks and morale signals by mining cross-repo traces and session histories to categorize findings into FEELING and PAIN items. This connects diverse signals across your repositories, helping humans make informed prioritization and recovery decisions for system-wide reviews.