Mahal Systems

Overview › Benchmark catalog

FormFactory

Fits data-entry and onboarding agents. Borrow the atomic (one field) versus episodic (whole form) scoring split.

What it asks the agent to do

  • fill an online form from a source document such as a resume
  • map each piece of user data to the right field across varied layouts

What to borrow from it

case templategrading design

Caveats

Form and field counts not in the abstract; repo URL not found; no model passed 5 percent at release so it gives little signal between agents yet.

Details

Workloadsweb browsing, workflow automation, documents
Interactionsingle turn
Environmentbrowser
End state checkedYes
Toolsweb form interface, PyAutoGUI actions
Gradingprogrammatic check, exact match
SizeNot published
Datasynthetic
Personal datanone
Licenseunknown
Harnessnone
Statusactive

Links