Roughly a dozen Early Cyrillic letters have no Greek model, and where their shapes came from is one of the standing questions of Slavic palaeography. Three positions are held.
The majority reading derives them from Glagolitic. The two alphabets share an inventory of sounds, an order and a set of names, and Glagolitic is the older; on this account whoever assembled Cyrillic kept the Greek uncial wherever it would serve and took the remaining signs from the alphabet already fitted to the language, simplifying them to sit on a Greek page.
A second reading looks to Semitic scripts, comparing zhivete, sha and tsy with Hebrew or Samaritan letters. The comparison has been made since the nineteenth century and rests partly on shape and partly on the argument that a Thessalonican scholar of the period could plausibly have known such letters.
A third position holds that neither derivation can be demonstrated. The shapes at issue are simple, and simple shapes resemble each other by accident; a resemblance is weak evidence in either direction.
What would settle it is a dated exemplar from the period when the choices were made, and none survives. The earliest manuscripts of both alphabets are later than the events they bear on. Until something older is found, or a decisive argument is made from the internal structure of the two letter sets, this stays a question with camps rather than an answer.