Intelligent Communication application programming interface (API) in a distributed system has become an integral part for developing smart home apps to connect intelligent computing hosts. The integration of intelligence into daily life of humans infusing smart behaviour in machines without human intervention or explicit knowledge is the dream of father of computing Alan Mathison Turing which he presented during the first half of twentieth century times and again. Turing test for a computer used to be a benchmark in this context. This smart connectivity within a home may assist humans to facilitate numerous varieties of connected appliances and may provide the efficient utilization of various apps to facilitate the end users. Intelligent homes with smart computing nodes may increase our productivity like machines did at the start of industrial revolution. The primary challenge for an API for intelligent homes is to make sure that the functionalities are provided in a trust-based network over the API which needs to be dependable, vigorous, and adaptive with the knowledge of the context around the distributed network so that the trust factor that comply to the specific needs of the inhabitants. In this research, the derivation and formal testing of trust computational model for Distributed Communication Services Programming Interface (DCSPI) for Smart Homes is presented. This computation is performed by developing actions through labelled transition system, behaviour and abstraction for each of the distributed components of the DCSPI. Each behavioural model component has intelligence of its context thus making it context-aware in terms of inter-process communication (IPC). Each Labelled Transition System (LTS) takes IPC equivalence computation from its concurrent LTS which reside in same state space to compute the trust weights between these. This IPC weighted-trust computation mechanism creates immunity in the DCSPI for external threats over the communication channel.