dsh-projects
Run the following command in DeepSeek Harness:
dsh plugin install alwaysmid-19990/dsh-projects
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install alwaysmid-19990/dsh-projects in your terminal, or download the plugin bundle manually from https://github.com/alwaysmid-19990/dsh-projects and install it locally; restart DeepSeek Harness Web to see the Projects entry in the sidebar.
About this plugin
Day-to-day work in DSH Web lives at the session level, with no persistent "project" concept to tie together a folder, dedicated instructions, and shared memory. dsh-projects brings a Cowork- and WorkBuddy-style project manager into your sidebar, giving you a full-screen panel to create, browse, and edit projects instead of juggling scattered sessions.
Each project is bound to a real working directory and carries its own editable instructions, a live file listing with manual pins, and an optional two-way CLAUDE.md sync: on save you can export instructions into the folder so that any AI tool (Claude Code, WorkBuddy, and others) working in that directory reads the same project context, or pull an existing CLAUDE.md back as the project instructions. All data is plain JSON stored locally with no hard-coded paths, making it trivially git-friendly.
If you use multiple AI coding tools and want one unified panel to manage project folders, instructions, and shared memory, dsh-projects was built for that workflow.
Use Cases
- Unify project context and instructions across multiple AI tools
- Maintain dedicated instructions, file lists, and pins per working directory
- Persist project folders and shared memory across sessions
Best For
- Using multiple AI coding tools like Claude Code and WorkBuddy simultaneously
- Wanting project-level instructions bound to real working directories
- Preferring local plain-JSON storage that can be tracked with git
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.