AI Agent Hub
Back to plugins
⚙️

dsh-rich-context

Workflow Updated 2026.08.29

Run the following command in DeepSeek Harness:

dsh plugin install svgop/dsh-rich-context

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install svgop/dsh-rich-context in your terminal to install the plugin. The source code is available at https://github.com/svgop/dsh-rich-context. After installation, restart the dsh web process and a Context entry will appear under Skill Center in the sidebar.

About this plugin

Anyone juggling multiple AI coding tools knows the drill: AGENTS.md or CLAUDE.md files are scattered across a dozen tool directories, and tweaking a single global instruction means dropping into a terminal, hunting for the right path, and editing blind in a plain text editor. dsh-rich-context pulls this workflow into the DeepSeek Harness sidebar. A new Context panel appears under Skill Center, featuring a monospace editor for both user-global and per-workspace instruction files, with Ctrl+S to save. The path bar in the header is click-to-edit, so you can route the editor to any absolute file path on the fly.

It also tames the cross-tool mess. A source scanner walks more than ten tool directories including .dsh, .codex, .claude, .cursor, .aider, and .gemini, listing every AGENTS.md and CLAUDE.md it finds with line counts. One click on Set as default symlinks the global config to the file you choose, making the harness transparently read whichever tool you designate as the single source of truth. Reset restores a plain file.

If you maintain instruction sets across several AI tools or want different projects to follow different agent behavior rules, this plugin moves config editing out of the terminal and into a purpose-built panel, cutting down the constant cat-and-vim round trips.

Use Cases

  • Editing global or per-workspace agent instructions frequently without leaving the web UI
  • Managing instruction files scattered across Claude, Codex, Cursor, and other tools from one place
  • Symlinking the harness config to any tool's AGENTS.md to establish a single source of truth

Best For

  • Power users maintaining instruction sets across multiple AI coding tools
  • Developers who prefer completing all configuration within the DeepSeek Harness UI
  • Small teams that need differentiated agent behavior rules per project