Quickstart

Prerequisites

As this guide utilizes a library published in Gen AI Google Cloud Repository, you will need to fulfill all the setup steps required in Gen AI's Prerequisites. To summarize, you need the following:

• Python 3.11 or higher

• Access to the GDP Labs' Gen AI SDK repository is required to run this program. If you need access, kindly submit a ticket to our DevOps team. Request access to Gen AI Google Cloud repositories by submitting this form (alternatively, from your manager or team lead)

• gcloud CLI: Please refer to the installation guide. After installing, please run gcloud auth login to authorize gcloud to access the Cloud Platform with Google user credentials.

• UV package manager

Getting Started

1

Installation

Initialize a new project and add the GL Connectors Tools package.

uv init --bare
uv add --extra-index-url "https://oauth2accesstoken:$(gcloud auth print-access-token)@glsdk.gdplabs.id/gen-ai-internal/simple/" gl-connectors-tools

2

Code

Create a main.py file with the following content:

main.py

from gl_connectors_tools.skills import SkillFactory
import asyncio

async def main():
    await SkillFactory.from_github(
        source="https://github.com/anthropics/skills/tree/main/skills/algorithmic-art",
        destination=[".agents/skills"],
    )

asyncio.run(main())

Parameters

Parameter	Description
source	Any GitHub path containing a SKILL.md file. Can be https, can be ssh.
destination	Local directory where the skill will be installed
token	(Optional) GitHub token for private repository access or for higher rate limits

Common Destination Paths

Agent	Path
General	.agents/skills
Claude Code	.claude/skills
Copilot	.github/skills
Cursor	.cursor/skills
Windsurf	.windsurf/skills
Codex	.codex/skills
Kimi Code	.kimi/skills

3

Run

Execute the script to install the skill:

uv run main.py

The skill will be downloaded and placed in your .agents/skills directory, ready for your agent to use.

4

What's Next?

You have the following options:

• You can implement your own agents that supports Skills! You can see Implementing with Agent for an example using LangChain's DeepAgents.

• You can now use the skills directly in your favorite Agentic Client such as Claude Code! For more information on how to install from private repositories, branches, etc. You can refer to the Sources to see what sources we currently support.