Skill UI

This workflow implements a bounded, iterative refinement loop for AI agents. It utilizes a required markdown rubric to grade the agent’s output in an isolated context. Based on the grader's verdict, it recommends the next move (sharpen, re-run, or promote). Once the agent passes the rubric, held-back evaluation cases are run in parallel to ensure generalization before final deployment.

This skill guides users through a structured, Socratic interview process to transform vague, high-level operational ideas into actionable, clear specifications. It incorporates advanced capabilities like lateral thinking review, ambiguity tracking, and iterative refinement, ensuring that all constraints and decisions are explicitly documented before development can begin.

Combines welfare readouts such as weight loss, temperature, clinical scores and biomarkers into a single RELSA severity score per animal per day, with kernel-density attention/danger zones and ARIMA-based forecasting of upcoming humane endpoints for 3Rs refinement reporting.