Rubisco is the main entry point of inorganic carbon into the biosphere and a central player in the global carbon system. The relatively low specific activity and tendency to accept O2 as a substrate have made Rubisco an attractive but challenging target for enzyme engineering. We have developed an enzyme engineering and screening platform for Rubisco using the model cyanobacterium Synechocystis sp. PCC 6803. Starting with the Form II Rubisco from Gallionella, we first show that the enzyme can replace the native Form I Rubisco in Synechocystis and that growth rates become sensitive to CO2 and O2 levels. We address the challenge of designing a zero-shot input library of the Gallionella Rubisco, without prior experimental knowledge, by coupling the phylogenetically guided model EV mutation with "in silico evolution". This multisite mutagenesis library of Synechocystis (n = 16) was subjected to competitive growth in different gas feeds coupled to deep sequencing, in order to compare Rubisco variants. We identified an amino acid exchange that increased the thermostability of Gallionella Rubisco and conveyed resilience to otherwise detrimental amino acid exchanges. The platform is a first step toward high-throughput screening of Rubisco variants in Synechocystis and creating optimized enzyme variants to accelerate the Calvin-Benson-Bassham cycle in cyanobacteria and possibly chloroplasts.