agency-evaluation-criteria

Evaluate AI agency deliverables by scoring them against weighted criteria like Design Quality, Originality, Completeness, and Functionality. This process uses input contracts such as BRIEF, Original copy, and design-spec files to generate a detailed evaluation report.

Yes, Playwright automates testing to verify deliverables against BRIEF, copy, and design-spec requirements. It captures screenshots and validates functionality to ensure AI agency outputs meet explicit pass/fail criteria.

An evaluation report includes an overall score, per-dimension scores with evidence, a defect list with file references and screenshots, and actionable improvement recommendations based on the input contracts.

Audit AI agency outputs by applying a standardized evaluation framework that scores originality and completeness against the original copy and design specifications, producing a defect list with file references for any missing elements.

You need to provide the BRIEF, Original copy.md, Original design-spec.md, and the built application artifacts. These inputs are used to evaluate the deliverables and generate the final evaluation-report.md.

Yes, the framework is applicable to agency projects including copywriting, UX/UI design, and implementation artifacts, enforcing quality gates across various deliverable types using a weighted scoring system.