harness-plan

To create a trackable plan for complex agent workflows, you can generate a standardized harness plan that defines the scope, size estimation, and acceptance criteria for future agents to execute without hidden context.

A self-contained plan for automated CLI execution is an auditable document that includes all required naming conventions, linting validations, and supplement documentation, ensuring agents can execute medium to large work without relying on chat history.

You need to use a standardized plan for agent execution when your project discovery has converged and you require a clean, auditable plan to guide future agents through medium or large work tasks without ad-hoc scripts.

To update an existing workflow plan with new scope, you apply scoped updates to the harness plan, running validation workflows and linting to ensure the modified plan remains executable and handles lightweight or heavyweight requirements correctly.

Yes, the plan creation process handles approval and linting workflows by enforcing validation steps that ensure the plan is approved, properly named, and size-estimated before it becomes an executable, auditable document.