Morphological analysis is a core component of natural language processing, particularly for mor- phologically rich and low-resource languages where data-driven approaches are often impractical. This paper presents a hybrid finite-state and rule-based morphological analyzer generator for Angika, an under-resourced Indo-Aryan language. The proposed system adopts a generator-oriented architecture that separates linguistic knowledge from core processing mechanisms, enabling modularity, extensibil- ity, and reuse across related languages. Finite-state transducers are employed to model regular inflec- tional morphology, while declarative linguistic rules handle irregular and constraint-sensitive phenomena that cannot be reliably captured through finite-state representations alone. The interaction between the two components is governed by a deterministic integration and conflict resolution strategy, ensuring predictable and interpretable analysis. Experimental evaluation demonstrates that the hybrid approach achieves higher coverage and accuracy than finite-state-only and rule-based-only baselines. The results confirm that a generator-based hybrid architecture provides an effective and scalable solution for mor- phological analysis in low-resource linguistic settings.