context-engineering

Curate codebase context by summarizing your stack and conventions, then converting project aspirations into observable rules using a HARNESS.md file. This makes guidelines explicit and enforceable for AI tools.

Context engineering structures the knowledge an LLM needs to work within a codebase by documenting the stack, defining conventions, and establishing an enforceability framework to keep guidance up to date.

Make coding conventions enforceable by defining observable rules for naming, structure, and error handling within a HARNESS.md file, ensuring properties can be checked by tools or reviewers.

Document stack details by declaring your current language, tooling, and environment within your project context. This explicit stack declaration ensures AI agents generate compatible code.

Maintain codebase context by applying maintenance guidelines that keep your HARNESS.md sections fresh and aligned with codebase changes across code, documentation, and review processes.

Convert aspirational coding guidelines into observable properties by utilizing reference materials and patterns that translate general goals into concrete, checkable rules for AI agents.