dsh-biomemory
生物仿生记忆系统（Biomimetic Memory for DeepSeek Harness）：纯文件 Markdown 数据层（透明可读改）+ memory 工具 + 冻结快照自动注入 + 分级审批门 + 结构化审计 + 记忆代谢（衰减/巩固/归档）+ 记忆钉 + 语义检索 + /memory 命令 + 跨会话召回
AI 분석
核心用途是构建长期、可代谢且透明可读写的个人知识库与记忆系统。适合需要跨会话长期记忆、信息沉淀与语义检索的复杂 Agent 任务。支持通过配置文件调整记忆半衰期等参数。
설치
$
npx -p @deepseek-ai/dsh dsh plugin --profile web add github:KLRSL/dsh-biomemoryConfiguration
// Plugin config (bundle or profile layer)
{
halfLifeDays: 7, // half-life in days for decay
decayThreshold: 3, // weight below this → archived
consolidateThreshold: 3, // references ≥ this → consolidate (+1 weight)
weightCap: 20, // consolidation weight cap (prevents runaway growth)
hotTokenLimit: 5000, // snapshot hot-section token budget
maxQueryResults: 20, // query result cap
petEndpoint: null // optional: local notification service URL (off by default)
}
Usage Scenarios
- Personal knowledge base, long-term maintenance: accumulate facts and decisions over time, query them later like a second brain; decay and archiving keep the store tidy without manual pruning.
- Project experience accumulation: lessons, habits and decisions live per-project in
projects//, consolidating (weight grows) as topics are referenced repeatedly. - Cross-session preference memory: preferences are injected at every session start, pin important ones for stability, and let conflict arbitration keep preferences authoritative over behavior.