dsh-plugin-workspace-rules
Run the following command in DeepSeek Harness:
dsh plugin install youjiaqi421/dsh-plugin-workspace-rules
Paste the following prompt into your AI chat to install this plugin:
Install the plugin in DeepSeek Harness using the command above; the source lives at https://github.com/youjiaqi421/dsh-plugin-workspace-rules and it automatically picks up Cursor, Gemini CLI, and Copilot workspace rules once activated.
About this plugin
Developers who use Cursor, Gemini CLI, and GitHub Copilot side by side tend to accumulate scattered rule files like .cursorrules, GEMINI.md, and copilot-instructions.md. When DeepSeek Harness enters the workflow, those instructions are easily left behind and the agent falls back to generic defaults. dsh-plugin-workspace-rules bridges that gap by pulling your existing workspace rules straight into Harness.
The plugin reads the three files in a fixed order, joins valid content into a single ordered system-prompt section, and injects it into every model request. The key design is that files are re-read before each request, so editing a rule file takes effect on the very next turn without restarting Harness. Missing or whitespace-only files are silently skipped; content is decoded as strict UTF-8 with a BOM stripped; per-file and total byte limits are enforced, and any violation fails loudly so the agent never runs with silently incomplete instructions.
It is built for developers who have already written rules for Cursor, Gemini CLI, or Copilot and want to add DeepSeek Harness to their daily flow. Open the plugin and reuse what you have already authored.
Use Cases
- A Cursor user wants to carry .cursorrules straight into a Harness workflow
- A team shares rule files across Cursor, Gemini CLI, and Copilot and wants them to take effect uniformly
- Editing a rule file and expecting the change to apply on the very next turn without restarting the agent
Best For
- Developers who use Cursor, Gemini CLI, or Copilot alongside DeepSeek Harness
- Workflow users who want to reuse existing rule files instead of rewriting them per agent
- Engineering teams that demand strict prompt completeness, UTF-8 validity, and replayability
Related Plugins
A method pack that makes AI coding agents plan against your real baseline, prove completion with fresh evidence, and reduce reworks and unsafe changes.
Turns the DeepSeek Harness session into a captain that builds a durable sub-agent team, splits goals into dependency-aware tasks, and coordinates work via direct messages and a live Web UI.
Gives coding agents design judgment, letting Claude Code, Cursor, and 70+ agents generate and iterate high-quality UI, presentations, and graphics right from the terminal.
Run the Pi ecosystem's plugins on DeepSeek Harness, unmodified, via a compatibility layer that implements Pi's public extension ABI on DSH's native services.