eval-harness

SKILL.md

Evaluation-driven development (EDD) is a methodology where evaluations are defined before or alongside implementation, ensuring that success criteria are explicit, measurable, and testable from the start.

Purpose: Verify that a new capability works correctly. Capability evals test whether the system can do something it couldn't do before, or does something better than before.

Purpose: Verify that existing functionality still works after changes. Regression evals protect against unintended breakage.

Comprehensive evaluation framework for systematic testing, measurement, and quality assurance of AI-assisted implementations. Supports capability evals, regression testing, multiple grader types, and standardized metrics. Source: mhylle/claude-skills-collection.

View raw