Skill Seekers: turn a GitHub repo into an Antigravity skill

This tutorial shows how the presenter uses Skill Seekers to turn a GitHub repository into a Claude-formatted skill package, then adapts the output for Antigravity. The example uses the Typer repository. The walkthrough covers installation, GitHub authentication, repository extraction and a prompt that tests whether Antigravity selects the generated skill.

The presenter starts with pip installation and configures a GitHub profile with a name and token. He explains that repository extraction can encounter rate limits and recommends a wait strategy with a 60-minute setting. That is his chosen configuration, rather than a guarantee that every repository will finish within that period.

In the demonstrated extraction, Skill Seekers uses the GitHub token and produces assets, references and a skill.md file. The presenter opens the output folder in Antigravity and asks it to turn skill.md into a skill stored in its own file system. He then asks how to create subcommands in Typer and reports that Antigravity detects and uses the skill.

The demonstrated example is Typer; the broader claim about converting any GitHub repository is not tested here.