Context is an important aspect towards service discovery and selection. It is represented by a set of quality parameters. Any change in value of any one of the context parameter’s (CP) changes the entire context. Relevance of the discovered services is often measured by similarity between service context and user’s context. If these two does not match for a particular user’s query; then corresponding services cannot be invoked or even if invoked, would perform poor. This paper proposes a novel context management framework. This holds the context information within a domain in a structured way such that the service discovery mechanism works faster as well as yields better result in terms of relevance of services specific to the queries from user. Autonomy, reactivity, and veracity properties of an agent help in achieving improved dynamics for the proposed framework. Implementation of the concepts and a comparative study is also reported. The proposed framework performs well with respect to search time, population size as well as varieties of queries.