dsh-layered-memory
Run the following command in DeepSeek Harness:
dsh plugin install DDDFXYqiming/dsh-layered-memory
Paste the following prompt into your AI chat to install this plugin:
In DeepSeek Harness, run the command `dsh plugin install DDDFXYqiming/dsh-layered-memory` to install the plugin, with the full source code available at https://github.com/DDDFXYqiming/dsh-layered-memory.
About this plugin
Managing long-term memory across sessions is a core challenge in AI, where the model needs to retain key facts, learn from errors, and retrieve information efficiently. The dsh-layered-memory plugin offers a structured solution for DeepSeek Harness, using a three-layer memory model (L1 index, L2 facts, L3 experience) and namespace isolation to systematize memory management, preventing information loss or chaos.
Its core capabilities include automatic index injection to maintain context, BM25 full-text search across namespaces, and mechanisms like content-level deduplication, archiving, and rollback to ensure data integrity and traceability. Guided by the "No Execution, No Memory" axiom, it only writes verified information, with tools like memory_write and memory_search providing fine-grained control for intelligent and reliable memory maintenance.
This plugin is ideal for DSH developers, researchers, or those handling complex tasks who require robust long-term memory support—such as managing project knowledge bases, debugging session histories, or building progressive learning systems. It streamlines memory management, enabling AI assistants to make more stable and efficient decisions based on historical information.
Use Cases
- Managing a long-term knowledge base for AI projects
- Tracing back historical session facts during debugging
- Building the memory layer for progressive learning systems
Best For
- DSH developers requiring persistent memory
- Researchers handling complex tasks
- Teams building knowledge-intensive AI applications
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.