What survives is modest almost to the point of invisibility: a rectangular structure built from drystone walling, a technique requiring no mortar, relying instead on the careful stacking and balancing of stones, and even that survives only to a height of about thirty centimetres, two courses at most. Half the building is simply gone.

Excavation carried out in 2001 and reported by Hayden in 2003 established what little can be known with confidence. The northern half of the house survived; the southern half did not. The walls, where they remained, were between 0.8 and 1.2 metres thick, substantial enough to suggest a building meant to endure. A gap of 1.2 metres in the east wall indicates where the entrance doorway once stood. No floor surface survived, and no occupation deposits remained, which means the daily textures of life inside the building, hearth ash, discarded bone, worn stone, left no trace. No datable finds were recovered either, so the late medieval attribution rests on the character of the structure itself rather than any object pulled from the soil. That places it loosely somewhere between the twelfth and sixteenth centuries, a broad window, but the only one available.

What makes the site quietly affecting is precisely its incompleteness. A building reduced to two courses of stone on a coastal headland, with no floor and no finds, offers almost nothing to interpret, yet someone chose this particular south-east-facing slope, probably for shelter from the Atlantic weather rolling in from the west, and built something there that lasted long enough to leave even this much behind.