agent-orchestration

把Claude Code的UltraCode模式带给DSH，把 DSH 的一次性多 Agent 调度，升级为可生成、可保存、可治理、可观察、可恢复的 Workflow 层

Persistent multi-agent rooms, shared timelines, and task coordination for DeepSeek Harness

Two-tier model router for DeepSeek Harness — LLM-Judge routing, multi-model fallback chains, exponential-backoff failover, and task-level orchestration (DSH adaptation of pi-shift-router)

Async LLM context relay between DeepSeek Harness sessions and Agent Team Rooms

Local-first issue board for DeepSeek Harness: board data, model-facing tools, a fresh-agent planning loop, and a human approval queue for agent-proposed issues.