constitution

Defining constitutional principles for an AI project involves a structured elicitation workflow that surfaces tensions across identity, autonomy, epistemics, and governance to prevent drift. This process produces operational principles and goals to guide development.

Documenting AI project governance is best achieved by generating governing artifacts like CONSTITUTION.md and GOALS.md. These files clarify behavior, establish operational principles, and document how decisions are enforced and reviewed.

To start a structured elicitation workflow, provide a project brief and run the elicitation process. This guides teams through crystallizing principles and boundaries, resulting in the generation of CONSTITUTION.md and GOALS.md files.

Revisiting governance and autonomy boundaries for an existing AI system is needed when restoring alignment. A structured constitutional elicitation helps surface tensions across identity and human-in-the-loop criteria to prevent project drift.

AI project governance documentation requires defining human-in-the-loop criteria. A structured elicitation workflow explicitly covers these criteria alongside actionability and epistemics to ensure decisions are enforced and reviewed properly.

Codifying goals and constitutional principles for autonomous AI systems produces two markdown artifacts: CONSTITUTION.md for operational principles and GOALS.md for personal objectives. These files guide development and clarify behavior.