by yoza10635

Guarded context compaction for DeepSeek Harness (dsh): the LLM proposes, deterministic guards dispose — eager per-atom shrink (extract/summary/false under verbatim guards) + lazy reference-graph evict

by fan56

dsh plugin: deterministic context compression backend — zero LLM calls, reproducible compression

Decision-and-goal memory with lineage: explicit commands and a model tool record decisions deterministically, a status card re-anchors them across compaction, a lossless event stream with BM25+jieba r

by MimicHunterZ

DSH plugin for agent-driven span compaction: compress chosen conversation spans into self-written checkpoints instead of the official head-anchored full-context sweep.

by skepsun

极简·省 token 的 DSH 长期记忆插件：零 LLM 自动捕获、符号索引渐进披露、ESR 证据闭环。Minimalist token-saving memory for DeepSeek Harness: zero-LLM auto-capture, symbolic index, ESR-lite evidence closure.

Fully local vector memory for DSH: local embeddings, SQLite storage, automatic recall injection, dedup with conflict detection, soft-delete recycle bin, online backup, and optional LLM extraction.