AI Agent Hub
Back to plugins
🧰

dsh-need-finder

Web Tools Updated 2026.08.17

Run the following command in DeepSeek Harness:

dsh plugin install 863683348/dsh-need-finder

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

Run dsh plugin install 863683348/dsh-need-finder in the DeepSeek Harness terminal to install this plugin; the source repository is available at https://github.com/863683348/dsh-need-finder

About this plugin

The DSH plugin catalog lists over a thousand entries by name and category, but your brain never thinks in plugin names. You think in tasks: notify me when a job finishes, view DSH remotely on my phone, scrape a webpage for the agent. dsh-need-finder bridges that gap. Tell the agent what you need in plain Chinese or English, and it returns the best-matching plugin from a curated set of 84, with a bilingual description, a matching rationale, and the install command. Order a dish, do not browse an aisle.

The matching engine is entirely local. No network calls, no LLM inference, no external dependencies. A weighted scoring algorithm hits on name (+3), tag (+2), and description (+1), with Chinese 2-gram substring matching and English tokenization, plus a category dictionary as a safety net. The catalog covers 14 categories across 84 real plugins, and the data file is plain JSON you can extend or trim on the fly.

Beyond single-plugin discovery, dsh-need-finder ships eight community recipe bundles (notification suite, security audit, remote mobile, dev tools, memory set, IM bridge, vision lab, research stack). One apply action outputs a dependency-ordered install plan that workflows and orchestrators can consume directly.

If you describe what you want to do rather than which plugin to install, dsh-need-finder is your starting point.

Use Cases

  • Describe a task in plain language and let the agent recommend the best-fit plugin with an install command
  • Apply a curated recipe bundle (notification suite, dev tools, remote mobile) to output a dependency-ordered install plan
  • Match needs across 14 categories from 84 curated plugins with zero network calls and zero LLM inference

Best For

  • DSH users who describe needs in natural language rather than plugin names
  • Developers who want a full plugin environment set up in one pass instead of browsing the catalog
  • New users who prefer semantic discovery over memorizing plugin names