Iteratively score and refine artifacts against rubric criteria until thresholds are met.
npx skills add https://github.com/sr-ai-dev/sr-harness --skill rulph
Please help me install this Agent Skill. Skill: rulph Source: https://github.com/sr-ai-dev/sr-harness/tree/main/skills/rulph Command: npx skills add https://github.com/sr-ai-dev/sr-harness --skill rulph
It prevents quality drift by turning subjective “looks good” reviews into an iterative rubric loop that can repeatedly score and improve an artifact until it passes a target threshold and per-criterion floor.
Use the rulph skill to iteratively grade and improve an artifact you care about by asking it to evaluate a file (or text) against criteria you provide, then let it loop until the required threshold and floors are met.
High-intent search queries and answers about installing and using this skill.
Rubric-based evaluation replaces subjective reviews with an iterative scoring loop, autonomously refining code artifacts against weighted criteria until a target quality threshold and per-criterion floors are met.
You set up iterative scoring by interactively defining evaluation criteria, locking a weighted scoring contract, and letting the system loop through autonomous improvements and rescoring until quality gates are satisfied.
Yes, multi-model parallel evaluation concurrently grades design artifacts using available models like Codex, Gemini, and Claude, aggregating their scores to ensure objective quality checks across all rubric criteria.
Criterion-by-criterion autonomous improvement repeatedly targets and refines only the single lowest-impact or floor-violating criterion, rescores the artifact, and continues until success or a circuit breaker stops the loop.
Circuit breaker stopping halts the iterative evaluation loop when progress stalls or maximum iterations are reached, preventing infinite cycles while preserving session state for safe Stop-hook integration and later resumption.
Yes, self-improvement relies on session state persistence to save scoring progress and artifact iterations, ensuring safe Stop-hook integration and allowing the autonomous improvement loop to resume without losing evaluation context.
Ankh-StudioEvaluate AI agent artifacts against a predefined rubric with weighted scores.
vigorX777Aggregate and score Hacker News blog articles into a Markdown digest.
jey3dayoManage iterative fix-and-review loops with parallel subagent investigation and quality gate checks.
o2b3kExecute iterative planning, production, evaluation, and refinement loops for complex artifacts.
