Coding Standards
In WW4, code formatting and coding standards are fist and foremost implemented and enforced by using objective code formation and code quality assessment by using a prescribed linter. For those of you not familiar with linters, this is the AI (Gemini) definition of a linter:
A linter is a tool that analyzes source code to identify potential errors, stylistic inconsistencies, and code smells. It helps developers improve code quality, maintainability, and security by flagging issues before they become problems in production.
For this purpose WW4 has adopted Clang (clang-format and cppcheck), which is integrated in the GitHub workflow by adding the file linting_cpp.yaml to the .github/workflow directory. When using Agentic AI for code development (e.g. Jules from Google, linting using Clang is also automated when using the linter settings in the AGENTS.md file in the root directory of the repository.
Final but important note: whereas we aim to automate most of this, we still enforce a human review of each pull request to the authoritative repository, which may result in additional formatting or linting requests. At NOAA, human review of code is a hard requirement when using Agentic AI for code development. Code generated using the AGENTS.md now automatically adds a statement to that effect to the documentation tags in the code.
