@vista8: Without good Skills, AI Agents cannot realize their potential. Another pain point is that with millions of Skills online, which one best matches my current task? Can it run? Is it safe? Vercel's find skill can search and install Skills, but only includes 90,000 Skills scattered across vertical domains…

AI Agents without good Skills cannot realize their potential. Another pain point is that with millions of Skills online, which one best matches my current task? Can it run? Is it safe? Vercel's find skill can search and install Skills, but only includes 90,000 Skills scattered across vertical domains, with little installation data to judge quality. Recently, I saw a Skill called SkillsVote from @MemOS_dev—it's quite impressive! It reportedly spent $200k using GPT5.4 to organize approximately 1.6M+ Skills from GitHub, filtering down to 790K+ Skills. Each Skill has its functional description, environment requirements, and runtime permissions extracted. On one hand, Skill recommendations are more precise; on the other, it records the execution steps of Skills and sends back feedback for subsequent optimization and recommendation. Best of all, it can even suggest workflow combinations based on the searched Skills, unlocking greater value from Skills. Vercel Skill only supports discovery and installation, while this Skill covers the entire cycle of "discovery → adaptation → attribution → iteration." ① It shows at a glance whether a Skill can run, what permissions it needs, and what packages it depends on. ② Whether execution succeeds or fails, and whether the Skill is poorly written or the environment is incompatible, the system can attribute and record it. ③ It generates evaluation tasks based on the Skill to verify if it meets the description. ④ As data accumulates, you won't step in the same pit twice. Website: https://skills.vote GitHub: https://github.com/MemTensor/skills-vote... This Skill supports Codex, Claude Code, 小龙虾 and other Agent clients. It's infrastructure built for AI Agents—recommend installing and trying it out.