Run existing skill evaluations via npx snapeval eval and parse benchmark results.
npx skills add https://github.com/matantsach/snapeval --skill run-evals
Please help me install this Agent Skill. Skill: run-evals Source: https://github.com/matantsach/snapeval/tree/main/skills/run-evals Command: npx skills add https://github.com/matantsach/snapeval --skill run-evals
Run and manage existing skill evaluations to accelerate iteration and improve skill quality.
Run the existing eval suite for this skill and report the results.
High-intent search queries and answers about installing and using this skill.
Run skill evaluations by invoking npx snapeval eval with your skill path. The process parses benchmark results, surfaces patterns like always-pass or always-fail cases, and proposes concrete improvements to accelerate iteration.
Re-evaluate iterations by running the evaluation suite after skill changes. The workflow parses benchmark results to measure impact, surfaces differentiating patterns, and proposes concrete improvements to optimize skill quality.
Yes, you can gate CI by running snapeval eval to generate benchmark results. The evaluation workflow orchestrates this process, parsing outcomes and surfacing patterns to validate skill performance before deployment.
Add or modify eval cases directly in evals.json while preserving existing ones. Control execution scope using the --only flag, then run snapeval eval to validate the updated suite and review benchmark results.
Yes, an existing evals/evals.json file is required. The evaluation workflow applies when the target skill already has evals defined, orchestrating npx snapeval eval to run benchmarks and surface performance patterns.
Review benchmark results for always-pass, always-fail, and differentiating patterns. Surfacing these patterns helps identify evaluation gaps and propose concrete improvements to optimize skill performance during iteration.
matantsachAnalyze a target SKILL.md and generate an initial evals.json with test scenarios.
Shockwave2kDraft SKILL.md files and run evaluation loops to measure trigger accuracy.
AminAlamAutomate test execution and gating in CI/CD pipelines for medical devices.
AlvaroBianoBenchmark language models on standardized tasks across API and local endpoints.
