DawnEver/cc-market

Iterative TDD review→fix loop — each round critiques the codebase, fans out fix agents over disjoint files, verifies, gates hard calls to the human, keeps tests green, and commits. Severity-based convergence with safety caps.

REM sleep for Claude sessions — auto-prunes stale memory on session start, crystallizes learnings on session stop, maintains .claude/memory/ with timestamps and 90-day eviction.

Post-feature sharp review — 3 parallel reviewers with schema enforcement, merge findings, sync structured task list with memory cross-references.

Multi-model AI orchestration — delegate tasks and planning to different model providers (DeepSeek, etc.) directly from Claude Code.

Local-first personal observability for Claude Code — token usage, cost, tool usage, daily reports. Hooks capture session/tool/prompt data, transcript parser extracts token/cost.

Unattended server & long-running task supervision — health checks, anomaly detection, auto-repair, multi-channel alerting. One YAML config per project, zero code changes to adopt.