Code Implementation

Overview

Code implementation standards require software to be consistent, maintainable, traceable, and secure.

Choose a standard below to read it in full.

Every codebase is held in one authoritative, traceable repository that excludes secrets from its history.

Every repository follows one defined branching model, with short-lived branches merged through reviewed and verified changes.

One style guide per language, applied consistently regardless of whether a person or an AI wrote the code.

Automated linting is applied consistently, with blocking violations resolved through the normal engineering process.

Static analysis continuously checks code for defects, security weaknesses, and maintainability issues, beyond style and formatting.

Every change is reviewed and approved by someone other than its author, with scrutiny proportionate to its risk.

Settings that vary by environment are externalised, and secrets are excluded from configuration.

Every dependency and runtime is deliberately selected, sourced through governed channels, and kept current.

Data access uses a defined layer, safe queries, and explicit, bounded transactions.

Every planned schema or data change is defined as version-controlled migration code and applied consistently through dedicated tooling.