DSH插件商店

dsh-agent-memory

findshan

Self-evolving memory for DeepSeek Harness: capture → dream consolidation → retrieval injection → evolve. User profile, project memory, correction learning, resume narrative — with provenance back to the replayable session log. 自进化记忆插件。

s1cap

Jaffe2718

System-1 decision models (Jev/Laya/Kev-class) as the governance layer for an LLM agent context lifecycle: growing association graph over session segments, relevance-gated recall with Trace-as-State ordering, and plan pre-ranking — measured in solve rate, cache hit/miss tokens, cost and latency. DSH plugin + harness-agnostic proxy.