中文Saved
KB Retriever
Project page on GitHub

SkillLift has not run this. The preview comes from the project, and the description is based on its README.

Open source

KB Retriever

Answers questions from a local folder of Markdown, PDF and Excel files, and says where each answer came from.

Garden Skills · ConardLi · MIT

Install

Inside Claude Code

/plugin marketplace add ConardLi/garden-skills
/plugin install knowledge-base-skills@garden-skills

Commands are copied word for word from the project's README. Read them before you run them.

Why it was picked

Four well-finished skills for motion, recorded talks, long reads and finding facts, each confirming the plan before it builds.

Why it exists

Agents answer from memory when they should check your files. This makes it read the folder first.

What you get

Answers with sources, found by reading only the parts of each file that matter.

How it works

  1. Add the plugin.
  2. Put your files in one folder, with a short index file as the project describes.
  3. Ask your question.

What it needs

Python with pdfplumber and pandas for PDF and Excel files.

In short

Pick this when the facts already exist somewhere in your files.

Worth knowing

SkillLift lists four of this collection's six skills. The other two are not listed: one offers style recipes named after real brands, and the other is built around a paid image model. The install command above adds only this skill.

Used in these workflows