What problem does it solve? Evaluating candidate material designs consistently across multiple expert perspectives is slow and prone to bias when a single reviewer covers every dimension. This Skill provides independent, rubric-driven scoring of material design candidates so each assessment is structured, reproducible, and free from cross-assessor contamination. ## Core Features & Use Cases - Independent Expert Perspectives: Loads one of three rubrics (catalytic activity, structural stability, environmental safety) per dispatch, with no shared results between assessors. - Structured Scoring Output: Produces JSON with per-dimension 1-10 scores, analysis, recommendations, and an overall score tied to the candidate's snapshot hash and version ID. - Weighted Aggregation Support: Feeds a server-computed final score (0.35 activity + 0.35 stability + 0.30 sustainability) for cross-validated candidate ranking. - Use Case: A lead research agent checkpoints a water treatment catalyst candidate and dispatches three independent assessors; each returns dimension-level scores and improvement suggestions that an aggregator synthesizes into a final ranking. ## Quick Start Evaluate this checkpointed material design candidate using the agent-a catalytic activity rubric and return the structured JSON assessment.
