harness-engineering

You can assess codebase readiness for AI coding agents by running a harnessability scan that measures type coverage, test coverage, architecture fitness, and documentation coverage to validate operational readiness.

Codebase harnessability for AI coding agents is a composite metric indicating whether the environment has sufficient type coverage, test coverage, architecture fitness, and documentation coverage to be safely operated by automated agents.

Yes, you need a TypeScript environment and npm installed to run the codebase harnessability scan, as the Skill requires npm to execute its validation scripts during the review phase.

You should run an AI agent readiness assessment during the validation and review phases, specifically before marking a sprint or epic done to confirm that codebase harnessability did not regress.

To execute a harnessability check using npm, run the command `npm run harness:scan` in your terminal to evaluate the codebase against type, test, architecture, and documentation coverage metrics.

The best way to prevent AI coding agent readiness regression is to integrate a harnessability scan into your validation and review phases to continuously track type, test, architecture, and documentation coverage before closing sprints.