I Published Skills to 5 AI Marketplaces in One Week — Here's the Brutal Automation Ceiling Test

In a recent effort, the author managed to publish agent skills to five AI marketplaces within a single week. The revenue generated so far is $0, but the real takeaway from this endeavor was the development of an "automation ceiling test." This framework serves as a decision-making tool before writing any code, helping to avoid unnecessary work. The test categorizes marketplaces into three tiers based on their publish paths, which are either fully automated, partially automated, or require human intervention at every step.

The author discovered that not all platforms are created equal when it comes to automation. For instance, Xiaping offered a full REST API, making the publishing process seamless with multipart upload and JSON arrays. Apify also provided scriptable publishing with its CLI tool, while MCPize and Capafy offered similar opportunities with CLI interfaces.

However, Agensi, despite having a more visually appealing interface and higher pricing, lacked a publish API and relied entirely on a web UI, making it unsuitable for automated workflows.

One key insight from the experience was the importance of thoroughly mapping the automation ceiling before investing time and effort into building a skill package. The author learned that each marketplace presents unique challenges, such as trial periods, promotion strategies, and distribution hurdles. While revenue share and payment models vary, the author emphasizes that automation is the most critical factor in determining the success of a published skill.

The author concludes by advocating for a repeatable, mostly-automated pipeline for building and publishing skills. By treating trial periods as a product phase, monitoring notifications, and actively engaging in community promotion, the chances of a skill gaining visibility and generating revenue increase significantly. The author encourages readers to share their experiences with automatable marketplaces, aiming to create a comprehensive table of automatable platforms for future reference.