dsh-memory-manager
Run the following command in DeepSeek Harness:
dsh plugin install tuanmaoOVO/dsh-memory-manager
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install tuanmaoOVO/dsh-memory-manager in your DSH environment; the source is at https://github.com/tuanmaoOVO/dsh-memory-manager. Restart DSH after installation to activate the plugin.
About this plugin
In DeepSeek Harness, an Agent starts every new session from a blank slate, forgetting prior debugging context and unable to consistently follow team conventions. dsh-memory-manager turns that amnesia into a durable, observable knowledge base by persisting memories as Obsidian-compatible Markdown files that outlive any single conversation.
The plugin revolves around a practical loop: each memory pairs an immutable snapshot with an editable annotation layer, tagged and toggleable to separate convention rules, session summaries, and general experience. When auto-injection is on, every new session automatically receives all enabled convention memories plus the latest eight session summaries, with no manual setup. Six built-in tools let the Agent search, save, enable, and plan memories on its own, while the browser side offers a three-tab panel for plans, library, and messages, plus a standalone force-directed knowledge graph with both semantic-panorama and focus-exploration views.
If you are building multi-session Agent workflows in DSH and want conventions applied without pasting them every time, a delegation-verification-retrospection loop that actually holds, or cross-project experience that is searchable and composable, this plugin was made for that scenario. It runs as a host-level plugin with on-disk storage, depends on no official plugin packages, and is designed so that a DSH upgrade degrades only individual features rather than breaking the plugin entirely.
Use Cases
- Keeping the agent aligned with team conventions across every new session without re-pasting them
- Reusing cross-project experience automatically in subsequent conversations
- Exploring memory relationships visually and jumping back to the source message in one click
Best For
- Developers building multi-session Agent workflows on DSH
- Engineering teams that rely on a consistent delegation-verification-retrospection loop
- Tech leads who want a markdown-first experience library for their agents
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.