add-vercel

SUSPICIOUS: the stated purpose is plausible and the Vercel CLI source is mostly coherent, but the skill’s core auth model routes a powerful Vercel token through third-party OneCLI infrastructure and then propagates it to all agents. That combination makes the data flow and permission scope disproportionate to a narrowly scoped deployment helper.

SUSPICIOUS. The core Vercel functionality is aligned and uses the official CLI, but the skill’s auth model routes credentials and API traffic through OneCLI’s third-party proxy using placeholder-token indirection. That intermediary credential/data path, plus raw OneCLI installer trust and delegated agent chaining, makes the footprint higher risk than a normal direct-to-Vercel deployment skill.