Evaluate agent outputs using multi-dimensional rubrics and weighted scoring.
npx skills add https://github.com/Asmayaseen/hackathon-2 --skill evaluation-asmayaseen
Please help me install this Agent Skill. Skill: evaluation Source: https://github.com/Asmayaseen/hackathon-2/tree/main/.claude/skills/evaluation Command: npx skills add https://github.com/Asmayaseen/hackathon-2 --skill evaluation-asmayaseen
The Evaluation skill provides a structured framework to systematically measure and compare agent outputs, enabling reproducible performance assessments across iterations.
Run the included EvaluationRunner against a predefined TestSet to generate a summarized report.
High-intent search queries and answers about installing and using this skill.
You can evaluate agent performance by applying a multi-dimensional rubric with criteria like factual_accuracy and tool_efficiency. This structured framework measures and compares outputs systematically across iterations.
A multi-dimensional rubric is a structured scoring framework using metrics like factual_accuracy, completeness, and citation_accuracy. It enables reproducible performance assessments by computing weighted totals across test datasets.
You can monitor production quality by using production monitoring hooks to sample and summarize agent performance. This tracks real-world behavior trends over time across live interactions.
Yes, you can track evaluation trends by running orchestration tooling against predefined test sets. This computes weighted scoring and generates summarized reports to compare performance across iterations.
To measure context engineering choices, you need a multi-dimensional rubric, test datasets, and tooling support. Running the included EvaluationRunner against a predefined TestSet generates a summarized report.
The best way to compare agent outputs is using a structured evaluation framework with weighted scoring. This approach enables reproducible performance assessments by systematically measuring results across test sets.
FVossebeldEvaluate agent outputs with weighted rubric criteria and per-dimension scores.
mhintz1980Evaluate agent systems with multidimensional rubrics and automated or human modes.
hussain9491Evaluate agent performance using structured rubrics and test sets.
ovachieverEvaluate scholarly work using the ScholarEval framework with dimension-based scoring.
