home / skills / partme-ai / full-stack-skills / skill-installer

skill-installer skill

This skill helps you discover, install, and manage AI skills from the PartMe marketplace, streamlining local skill deployment for apps.

npx playbooks add skill partme-ai/full-stack-skills --skill skill-installer

Review the files below or copy the command above to add this skill to your agents.

Files (9)

SKILL.md

1.1 KB

---
name: skill-installer
description: Manages the installation and discovery of AI skills from the PartMe marketplace.
---

# Skill Installer

## Overview

The `skill-installer` is a core utility skill that allows users and agents to discover, install, and manage other AI skills within the PartMe ecosystem. It serves as the local package manager for the "Knowledge as a Service" (KaaS) and "Tool as a Service" (TaaS) architecture.

## Features

- **Search Skills**: Query the local marketplace for available skills by keyword.
- **Install Skills**: Register skills into the local environment.
- **List Installed Skills**: View currently active skills.

## Usage

### Search for a skill
```
skill-installer(action="search_skills", query="vue")
```

### Install a skill
```
skill-installer(action="install_skill", skill_path="./skills/vue3")
```

## Tools

### `search_skills`
- **Input**: `query` (string)
- **Output**: JSON list of matching skills with descriptions and paths.

### `install_skill`
- **Input**: `skill_path` (string)
- **Output**: Confirmation of installation.

### `list_installed_skills`
- **Input**: None
- **Output**: JSON list of installed skills.

Overview

This skill manages discovery, installation, and basic lifecycle actions for AI skills from the PartMe marketplace. It functions as a local package manager for skills used by agents, enabling quick search, install, and listing of installed capabilities. It streamlines integrating new tools into agent workflows with minimal manual steps.

How this skill works

The skill inspects a local marketplace index and returns matching skill metadata when queried. Installing a skill registers its path and metadata into the local environment so agents can load it at runtime. A listing operation returns the current set of installed skills and their basic descriptions for auditing or selection.

When to use it

You need to add new agent capabilities from the PartMe marketplace.
You want to discover available skills by keyword before installing.
You must audit or list which skills are currently installed locally.
You want a repeatable, scriptable way to register skills in an environment.
You need to automate skill installation as part of deployment pipelines.

Best practices

Search with focused keywords and review described paths before installing.
Install skills from trusted sources and validate their metadata and version.
Use consistent directories for skill_path values to simplify environment management.
Keep a record of installed skills and their versions to enable reproducible builds.
Test newly installed skills in an isolated environment before promoting to production.

Example use cases

Quickly find a frontend helper skill: search for "vue" to get matching skill entries and paths.
Programmatic installation in CI: call install_skill with a repo or local path to register capabilities during build.
Inventory check: run list_installed_skills to generate a JSON audit of active skills before deployment.
Agent onboarding: an agent queries the marketplace and installs required skills on first run to adapt to task needs.
Skill updates workflow: uninstall or overwrite a skill path, then reinstall a newer version and confirm via listing.

FAQ

What inputs does the installer accept for search and install?

Search accepts a query string. Install accepts a skill_path string pointing to the skill directory or package location.

What does listing return?

Listing returns a JSON array of installed skills with basic metadata such as name, description, and registered path.

Can I automate installs in scripts or CI?

Yes. The installer is designed for programmatic use; call the install action with a path in scripts or pipelines.