Agent examplesPolicy research
Program evaluation
Goals from the program's own documents, each effect estimated against places it did not reach, revenue against projection, every goal rated. Every figure traced to its source
The problem
Goal met on paper: 11 percent fewer vehicles against a 10 percent target
The MTA's 73,000 fewer vehicles a day, 11 percent, sets zone entries against traffic before the toll. That count also carries everything else that moved traffic since January 2025: return to office, construction, weather, the economy. Grade the 10 percent objective on it and the program passes or fails on the wrong number
The request
How Axe works
What you get back
Make it yours
Point it at another program, set your own goals and rating thresholds, or change the comparison groups and the period. Every step and every prompt can be edited
Opens a private copy of this Agent in your workspace, ready to adapt to your own sources