Perhaps agency is less one single 'meta-skill' than a bunch of domain specific 'micro-skills'. E.g., strategic thinking in a tech company leadership roll is one skill (or set of skills), strategic thinking in the context of a personal relationship is another skill (or set of skills); managing fear in the context of public speaking is one skill (or set of skills), managing fear in the context of driving is another skill (or set of skills).
Sometimes skills that seem superficially similar are surprisingly non-transferable and context specific.
Ultimately, I think that the most generally applicable advice is getting feedback from others, getting an outside view, blindspot fuzzing, etc. Bottlenecks are usually not the obvious things, because usually the obvious things have been tried, without success. [That is, until the obvious thing is so obvious that everyone assumes it has been tried before without actually trying it. In which case it is again non-obvious!]
Personally, I have found that in personal relationships, a fairly effective strategy to get "unstuck" is to take actions that seem bizarre, unimaginable, or otherwise far outside the distribution of my ordinary responses. I do this because usually if I feel stuck, or there is some unpleasant dynamic, then a part of that dynamic that keeps it going is my actions, which are based in what I tend to do. If I find myself in the 3rd, 4th, or 5th iteration of something unpleasant that I feel like has played out before, I'll do something far outside the distribution of my ordinary actions, and it has a much better chance of getting me unstuck (though, it still might only be a 1%-10% chance of getting me unstuck per random action).
Perhaps there is some sort of general algorithm in that for getting unstuck. A sort of exhaustive process of elimination, with some sort of outside generativity, and some feedback loops grounded in reality, that is one's best chance at escaping local optima.
Sometimes skills that seem superficially similar are surprisingly non-transferable and context specific.
Ultimately, I think that the most generally applicable advice is getting feedback from others, getting an outside view, blindspot fuzzing, etc. Bottlenecks are usually not the obvious things, because usually the obvious things have been tried, without success. [That is, until the obvious thing is so obvious that everyone assumes it has been tried before without actually trying it. In which case it is again non-obvious!]
Personally, I have found that in personal relationships, a fairly effective strategy to get "unstuck" is to take actions that seem bizarre, unimaginable, or otherwise far outside the distribution of my ordinary responses. I do this because usually if I feel stuck, or there is some unpleasant dynamic, then a part of that dynamic that keeps it going is my actions, which are based in what I tend to do. If I find myself in the 3rd, 4th, or 5th iteration of something unpleasant that I feel like has played out before, I'll do something far outside the distribution of my ordinary actions, and it has a much better chance of getting me unstuck (though, it still might only be a 1%-10% chance of getting me unstuck per random action).
Perhaps there is some sort of general algorithm in that for getting unstuck. A sort of exhaustive process of elimination, with some sort of outside generativity, and some feedback loops grounded in reality, that is one's best chance at escaping local optima.