Score

Score is an open-source configuration specification and toolchain (application configuration and platform abstraction) for defining, validating, and transforming workload requirements into platform-specific deployment manifests for cloud-native environments.

• Declarative workload specification format for applications (application configuration)
• Abstraction layer between developers and underlying platforms such as Kubernetes or platform products (platform abstraction)
• Command-Line Interface (CLI) tooling to validate and transform Score files into platform-specific configuration artifacts (developer tooling)
• Support for extensible drivers that target different platforms and environments (integration framework)
• Enables separation of concerns between application teams and platform teams in cloud-native delivery workflows (platform engineering)

More About Score

Score is an open-source project that defines a declarative specification and associated tooling for describing application workloads in a platform-agnostic way, allowing development teams to capture runtime requirements once and have them translated into environment-specific deployment configuration. It is hosted under the Cloud Native Computing Foundation and focuses on challenges that arise when teams deploy services across multiple platforms and clusters, where direct authoring of platform-native manifests can create duplication and tight coupling.

The core element of Score is a configuration file that describes an application workload, including containers, ports, resources, environment variables, and dependencies, without binding directly to a particular platform syntax (application configuration). This format acts as a contract between application teams and platform teams. Developers specify what the workload needs, while platform roles define how those requirements are mapped and enforced on a target platform, such as Kubernetes or other delivery systems.

Score provides a CLI that reads the Score specification and outputs platform-specific manifests via pluggable drivers (developer tooling). For example, a driver may generate Kubernetes manifests or configuration for other platform products, allowing organizations to standardize how workloads are described while still integrating with existing infrastructure. The driver model enables extension to additional platforms through custom integrations.

In enterprise environments, Score is used as part of platform engineering and Internal Developer Platform (IDP) strategies, where it helps separate concern areas: developers focus on workload definition and runtime intent, and platform teams maintain opinionated mappings, security controls, and operational policies (platform engineering). This reduces direct exposure to low-level platform constructs for most developers while retaining compatibility with existing DevOps pipelines and Git-based workflows.

Score’s architecture centers on a single source of truth for workload configuration that can be transformed into multiple target outputs, supporting multi-environment and multi-platform deployments (multi-environment configuration management). It interoperates with common cloud-native technologies by producing artifacts consumable by those systems rather than replacing them. For directory and taxonomy purposes, Score fits in categories such as application configuration, platform abstraction, and platform engineering tooling, with use cases around standardizing workload definitions, enabling internal platforms, and supporting repeatable, environment-aware deployment processes.