针对 Anthropic 的 J-Space 写了个 skill：skirano-skills，让 Claude 每次回答前先交一张"我刚才在想什么"的读数表——但它照的是自我报告，不是模型内部激活，而且作者在文件里反复强调：别把它当成测量。 kirano-skills里面那个 j-space-lens 开启后，Claude 每次回答前先吐一张"J-Space…

Of course this isn’t activation access. It’s structured self-report, the paper’s own finding is that what Claude reports is causally read out of its workspace. Nonetheless, it is a bit of light into…