Modern networks are operated through vendor-specific tools, manual configuration, and tribal knowledge — an approach that does not scale and leads to configuration drift and slow fault response. This paper proposes a declarative, vendor-agnostic Network Intent Schema: a hierarchical JSON model capturing network design at three levels (network, node, and link). The schema serves as a single source of truth driving six operational use cases: configuration generation, rack elevation, port mapping, bill-of-materials creation, state validation, and topology-aware fault detection. A reference implementation covers a three-tier campus network, a four-router WAN backbone, and a VXLAN/eBGP EVPN datacenter fabric across two sites. A brownfield bootstrap strategy using LLDP discovery and AI-assisted intent completion makes adoption practical for existing networks. This paper is the first in a series on self-healing network operations. Related work: Ghosh (2026), Network Health Score Framework (https://doi.org/10.5281/zenodo.20552168).