Developer open-sources 27 AI testing-specific agent skills

On November 5, 2026 at 09:21, Juejin published an article introducing 27 self-developed, open source Agent Skills built specifically for AI testing. According to the article, the skills cover the full workflow of requirement breakdown, test case design, and interface plus UI automation. All of them were built on real testing scenarios, can be used standalone or chained together in orchestration, and include capabilities such as converting requirements into user stories and generating test cases. The stated goal is to improve efficiency in real testing scenarios.

About one minute later, at 09:22 the same day, cnblogs carried a post from a blogger sharing the 27 AI testing skills they use day to day. The blogger said their skills span several directions, including office productivity, content creation, and automated workflows, and that this post would single out the ones most strongly tied to AI testing. More detail was promised in future posts, and the blogger noted that a comprehensive roundup of AI testing skills had already been published two months earlier.

The two sources agree on the number, 27 skills, but describe them differently. The Juejin piece stresses that the skills are self-developed and open source, that they cover the end-to-end testing workflow, and that they support chained orchestration. The cnblogs post frames them as the blogger's personal daily toolkit, touching a broader set of scenarios such as office productivity and automated workflows. Neither source provides a complete list of all 27 skills. The matter currently sits at the sharing-and-teaser stage: the skills have been introduced, and further write-ups are promised but have not yet appeared in the available material.