dsh-plugin-vault-memory
Run the following command in DeepSeek Harness:
dsh plugin install zhaoxuejie/dsh-plugin-vault-memory
Paste the following prompt into your AI chat to install this plugin:
Please install the dsh-plugin-vault-memory plugin in my DeepSeek Harness (source: https://github.com/zhaoxuejie/dsh-plugin-vault-memory), then configure my Obsidian vault path in the plugin settings and restart dsh.
About this plugin
Many users have built large Obsidian vaults, yet every new DeepSeek Harness session starts with zero context about your library structure, active projects, or past conclusions. Notes pile up but stay undiscoverable, session outputs scatter into chat history, and the vault grows messier over time. dsh-plugin-vault-memory turns your notes from a static archive into a live asset the agent can actually query.
In conversation, pull up old notes with natural language: full-text search with Chinese tokenization, structured queries by folder, tag, or date, and optional local semantic search. Every result carries the source file path, and the agent will say plainly when nothing is found rather than fabricating. Each new session auto-injects a memory snapshot (vault stats, active projects, recent notes) so the AI jumps straight to the task. Finished output saves as a note in one click with auto-generated frontmatter, tag suggestions, and related-note links. A daily health check flags orphaned notes, broken links, and missing index pages with reasons; nothing is written until you approve, and a backup is taken before any change.
All indexing, search, and inspection run locally; note content stays on your machine by default. Semantic search uses local Ollama (bge-m3) with optional cloud endpoints. This plugin is ideal for users who manage knowledge in Obsidian and rely on DeepSeek Harness for daily Q&A and output, especially those who want their agent to genuinely understand the note library.
Use Cases
- Ask the agent for prior research conclusions and get sourced answers from your vault
- New sessions auto-inject a memory snapshot so the AI dives straight into the task
- Save session output as a note with auto-tagging and related-note suggestions in one click
Best For
- Knowledge workers who use Obsidian for personal knowledge management and rely on AI agents daily
- Power users who want their agent to read their vault instead of starting from scratch each time
- Privacy-first professionals who require note content to stay on local machines
Related Plugins
Memory layer for coding agents that indexes local session history and auto-recalls relevant context before edits, commands, and failures, with no manual search needed.
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.