What problem does it solve? Running autonomous multi-epic implementation loops requires correctly generating scripts, launching them in detached terminal sessions, and continuously monitoring progress, stalls, and crashes — a process that is error-prone when done manually. ## Core Features & Use Cases - Script Generation: Generate infinity loop and polish loop scripts via the ca loop and ca polish CLI commands with flags for models, reviewer fleets, retry limits, and review cycles. - Detached Launching: Launch loops in GNU screen sessions (or tmux under WSL2 on Windows) with chained pipeline support for running infinity and polish loops sequentially. - Health Monitoring: Structured health-check protocol reading .loop-status.json, loop-execution.jsonl, and trace logs to detect stalls, crashes, retries, and human-required blockers, with per-epic progress tables and ETA estimates. - Use Case: After planning a set of beads epics, generate an infinity loop script, dry-run it, launch it in a screen session, then ask for a progress report showing completed epics, current epic, and anomalies. ## Quick Start Ask the assistant to generate and launch an infinity loop for your open beads epics and then report on its progress.
