On- and off-chain partitioning is a common challenge in blockchain applications. Once information enters a blockchain environment, the nature and architecture of a blockchain render this information trusted in terms of integrity. Analogously, information processing—or software execution in general—inside a blockchain environment (on-chain) is considered trustworthy. However, at some point, on-chain information has to rely on external information or perform software execution outside the blockchain (off-chain). Reasons can be, e.g., access to external data such as sensor values, preserving data privacy, specialized software execution capabilities of external systems or resource constraints of the on-chain environment. Proof-of-Existence (PoE) is a common application for blockchain technology. Research data certification is a particular application of ch1PoE, where researchers can attest data provenance by storing a hash value of their data on a blockchain. This application poses several challenges of on- and off-chain partitioning: it usually is desired that the research data stays private until its complete publication and the calculation of the data’s hash might utilize a high amount of resources when done on-chain. But the hash may be tampered with. Ideally, a blockchain would push a controlled software to generate the hash value at the researcher’s computer, and return the result as trusted on-chain information. This paper introduces, explores, and demonstrates the moving smart contract following this idealized process. In this concept, a smart contract distributes software accessed via the blockchain and manages execution results to include only confirmed execution results in the blockchain.