Turns out you can just get agents to do some apocryphal vulnerabilities research. Here’s two cuts from the output:

• Background. Many fields rest part of their authority on a founding anecdote, quote, statistic or single study. When that story proves false, the larger construct built on it may be exposed. We call this an apocryphal vulnerability, and set out to estimate how often such stories are load-bearing, how they arise, and what their failure costs.
• Conclusions. Apocryphal anecdotes are common across fields, but their danger depends on how much they carry. The highest-risk configuration is a vivid founding story, introduced by an interested party or a single weak study, that maps directly onto professional practice. Because correction typically takes decades and has accelerated only recently, fields with such stories at their base would benefit from a deliberate audit of their foundations.