dsh-motion-memory
Run the following command in DeepSeek Harness:
dsh plugin install li3-feng2-jie2/dsh-motion-memory
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install li3-feng2-jie2/dsh-motion-memory in your DeepSeek Harness terminal and restart DSH; you can also clone from https://github.com/li3-feng2-jie2/dsh-motion-memory and place it into your profile manually.
About this plugin
Every new session means the AI forgets everything — prior decisions, lessons learned, project context. That is the most frustrating limitation of conversational tools. dsh-motion-memory solves this by providing a full local memory-management layer for DeepSeek Harness: it automatically distills conversation content worth keeping into structured memory documents, so the AI genuinely gets to know you over time and you never have to re-explain your background.
Core capabilities include a four-tier memory structure (active, periodic, event, raw dialog) with full traceability down to session and turn level; per-turn dialog tracking that compresses and accumulates content automatically; periodic summarization that consolidates and compresses entries to prevent unbounded growth; keyword deduplication routing that updates the same entity and disambiguates new ones; and inline reference links that jump straight back to the source conversation. Everything runs locally with a small local model (7B to 9B is plenty), so no paid API calls are needed and your data stays fully under your control.
It is built for anyone who works long-term with DeepSeek Harness — developers, writers, researchers, or daily-assistant users who want the AI to remember project progress, personal preferences, and working habits, reduce repetitive context-setting, and gradually build a personal knowledge base. Contributors who want to help polish the memory pipeline are also welcome.
Use Cases
- Long-term project development where the AI remembers architectural decisions and progress without re-briefing
- Daily writing or research where the AI retains your style, terminology, and preferences over time
- Complex multi-turn conversations that automatically accumulate lessons learned, reducing redundant context-setting
Best For
- Power users of DeepSeek Harness doing daily development, writing, or research
- AI users who want to minimize repetitive context and build a personal knowledge loop over time
- Privacy-conscious users who prefer fully local, self-controlled workflows without paid API dependency
Related Plugins
Traceable, searchable cross-session memory for AI agents that turns conversation knowledge into a typed knowledge graph and recalls relevant subgraphs instead of replaying full history, natively integrated with DeepSeek Harness.
Gives DSH AI cross-session long-term memory, to-do and skill management, plus multi-session orchestration, external AI delegation, and an infinite canvas that grows with you.
dsh-mnemon is a three-tier, pluggable, Agent-driven memory system for DeepSeek Harness, combining Runtime memory, Project Documents, and replaceable Memory Spaces with nine long-term providers.
An opinionated, zero-infrastructure file-based memory protocol: human-readable markdown files plus a bounded always-loaded index, curation discipline, and no database or embeddings.