AI Agent Hub
Back to plugins
🧰

dsh-pkg-info

Web Tools Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install ZhijiangTang/dsh-pkg-info

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

Run dsh plugin install ZhijiangTang/dsh-pkg-info in your terminal to install the plugin. The source repository is available at https://github.com/ZhijiangTang/dsh-pkg-info.

About this plugin

In LLM Agent or Code Mode pipelines, models often need to confirm package metadata such as version, license, and dependency lists from npm or PyPI, but calling each registry HTTP API directly means dealing with inconsistent field names, scattered error branches, and return structures that differ per registry. dsh-pkg-info collapses all of this into a single pkg_info tool call: pass a registry, package name, and optional version, and you get back a structured JSON object with fully deterministic field names and types, removing the need for defensive parsing downstream.

The core design philosophy is determinism: on success the response carries ok as true plus normalized description, license, author, homepage, repository, version count, and dependencies; on any failure, whether a 404, network timeout, or JSON parse error, the response is folded into ok as false with a stage and message, so no exception ever escapes the tool boundary. The plugin has zero runtime dependencies, is pure ESM, and requires no build step, keeping integration costs minimal.

It is well suited for developers building Agent workflows or Code Mode tasks that need programmatic package lookup, and also serves as a lightweight metadata probe in RAG or code-generation pipelines, giving models structured package context at minimal integration cost.

Use Cases

  • Query package version, license, and dependencies within an Agent workflow
  • Retrieve normalized package metadata JSON programmatically in Code Mode
  • Supply structured package context to models in RAG or code-generation pipelines

Best For

  • Developers building LLM Agent workflows
  • Engineers needing programmatic package lookup in Code Mode
  • Teams building RAG or code-generation pipelines