dsh-skill-mcp-manager
Run the following command in DeepSeek Harness:
dsh plugin install alone-tree/dsh-skill-mcp-manager
Paste the following prompt into your AI chat to install this plugin:
To install this plugin in DeepSeek Harness, please execute the command `dsh plugin install alone-tree/dsh-skill-mcp-manager` in your terminal. The source code is available at https://github.com/alone-tree/dsh-skill-mcp-manager.
About this plugin
As SKILLs and MCP tools accumulate in DeepSeek Harness, managing them becomes a challenge, leading to context clutter and wasted tokens. dsh-skill-mcp-manager is designed to solve this, offering a one-stop visual interface that lets you clearly see all installed tools and their status without digging through layers of configuration.
The plugin's core strength lies in its "on-demand loading" mechanism. Unlike traditional methods that inject full schemas upfront, it only exposes tool names and short descriptions at the start of a session. This ensures a perfect balance between utility and cost, significantly reducing token usage while keeping the AI informed. It also features hot-reloading and session isolation, making it extremely developer-friendly.
Whether you are a developer needing multiple concurrent browser instances or a user seeking fine-grained control over tool invocation, this plugin is the ideal solution. It safely takes over your existing MCP configuration and prevents state conflicts across sessions. You can now focus on designing your workflow logic, leaving the complex tool management to the plugin.
Screenshots
Use Cases
- When managing many MCP tools and optimizing token usage.
- When running multiple browser automation tasks concurrently in one Harness instance.
- When wanting to hot-reload local MCP servers without restarting the plugin.
Best For
- Developers focusing on building complex DeepSeek workflows.
- Advanced users needing fine-grained control over tool invocation and session isolation.
- Engineers developing and debugging MCP servers.
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.