メモリ・ナレッジ · DeepSeek Harness プラグイン — 38 ページ

DSH plugin: dev progress tracking, system bug memory, token optimization, and context compact continuation

DeepSeek Harness（Cordis）插件：把 lhdrc-agent-memory（df-memory 核心仓）的记忆能力接入 agent 工具面。

DSH plugin: show a git-style file diff right inside the chat whenever the AI calls write/edit tools. Unified (top-bottom) and split (side-by-side) views with inline word highlighting, view-preference memory, zero git dependency.

C-drive planner for DeepSeek Harness: scan the whole C drive, classify every big item into A/B/C/D/E/P (safe cache / relocatable / junction-movable / abandoned software / config-memory redline / personal files), LLM fallback + human confirm, then delete/move/redirect on demand with strong redline + dry-run/undo safeguards. C盘空间规划整理大师:扫描整 C 盘,分门别类给出正确动作,动手前先预览,红线配置/记忆永不自动执行。

Zero-knowledge WebSocket relay for dsh-dispatch: routes end-to-end encrypted frames between DeepSeek Harness (dsh) machines and phones, and fans out Web Push when the phone is offline.

DSH host plugin: scheduled background 'dream' (memory consolidation) passes that distill recent session work into long-term memory (memory/MEMORY.md) and a dream diary (memory/dreams.md), without manual triggering. Includes dream_status / dream_run agent tools.

DSH 自动做梦插件：后台定期回顾近期会话与记忆，整合去重、更新过时、修剪索引（复刻 CC-HAHA autoDream，与 dsh-memory 并存）

dsh 梦境记忆整合插件：凌晨随机触发，把当天会话/Hindsight 记忆生成叙事化梦境日记（法塔人格），高价值洞察晋升回 MEMORY.md，梦境与晋升记录存 SQLite 并在 web 梦境 Tab 展示。

DSH Web 用户消息气泡内联编辑（rewrite）与撤回插件：点击气泡原位编辑、行内确认胶囊撤回、惰性提交、< X > 版本翻页器。

DSH 专用记忆插件：跨会话持久记忆（存储领域 + agent 记忆工具 + 系统提示词注入 + AI 建议记忆 + 设置面板卡片）

Host-only DSH plugin: remember the reasoning effort that was last in effect per (provider, model) and restore it when a session switches back

Minimalist long-term memory for DeepSeek Harness, distilling symbolic-index + pi-esr ideas: zero-LLM auto-capture, a symbolic [ENGRAM] index with progressive disclosure, and an ESR-lite evidence-closure protocol (esr_task / esr_node / esr_close / esr_link). Real agent-behaviour telemetry (usage rollup + /stats) and a deterministic offline recall benchmark (npm run eval). Hot path is model-free; storage on ctx.storageDomain; web memory viewer, observability panel and config card through DSH's native settings slots.

Per-session Engram memory for DeepSeek Harness: spawns an engram MCP child per agent session rooted at the session workspace, registers mem_* tools per agent scope, and injects the Memory Protocol as a system-prompt section.

Bundled evidence-first knowledge work skill for DeepSeek Harness

进化核心插件（心脏）：聚合全量进化器官（self-test/emotion/reflection/life-core/evolve/memory/checkpoint/skill-forge）实时状态 → 五环完整性诊断（猜想→采证→finding→裁决→布线）→ 断点识别 + 行动建议。感知圈驱动 + 信号浮出，决策归爱丽丝（只读取证、不自动执行）。

Self-evolving memory + skill lifecycle for DeepSeek Harness. Cross-session memory with zero-token deterministic recall (bigram-Jaccard fused with FTS5 BM25 via RRF), a tiered approval gate, and reinforcement that strengthens what you repeat. Procedural knowledge crystallizes into SKILL.md files that refine in place and are curated through an active-stale-archived lifecycle (reversible archive, pre-op backups, rollback, never deletes). Includes background per-turn review, anti-bloat convergence for both skills and memory, an auto-grown user profile, and a web settings page.

Auditable, verifier-gated self-evolution for DeepSeek Harness.

File-based self-evolving memory for DeepSeek Harness v2: compression & extraction over official compaction, daily episodic layer, dream consolidation to semantic files (user/agent/memory/dream + project), catalog disclosure injection, no confidence machine. 文件化自进化记忆：消费官方 compact 压缩提取，每日情景层 + 梦境整合固化语义，目录式披露注入，无置信度体系。

dsh-experience-flywheel · DSH 经验飞轮插件 — agent cross-session experience memory + dual-acceptance gates (auto query/inject via agent/pre-step middleware; exit-1 gate scripts)

Evidence-governed experience memory for DeepSeek Harness

Fact-checking skill for DeepSeek Harness: verifies claims against the open web instead of model memory, cross-checks independent sources across domestic and international search ecosystems, and can route three factual judgements (does the evidence support the claim / is this source primary / does this page carry usable text) to a TypeSafe Jev (System One) decision model when a key is configured, falling back to the skill's deterministic rules when it is not.

Fount-inspired, in-turn memory for DeepSeek Harness

Game-style floating HUD for DeepSeek Harness: HP bar (account balance vs ¥20 cap), MP bar (context remaining), official peak/valley pricing with switch countdown, auto-compaction when context < 5% remains, and a memory-carrying new-conversation button. · 游戏风格悬浮 HUD：余额血条、上下文蓝条、官方峰谷定价与切换倒计时、上下文不足5%自动压缩、携带记忆开启新对话。

Cordis Host integration for Engram persistent memory in DeepSeek Harness