#
ENJA

waza-orig

spboyer/waza-orig

Waza is a Go-based CLI tool and evaluation framework designed to benchmark, test, and assess AI agent skills and models.

7 4No licenseUpdated 2026-02-28

Overview

Waza provides a structured environment for AI agent skill development, featuring tools to initialize projects, scaffold evaluation suites, execute benchmarks against isolated workspaces, and validate outputs using configurable graders. It supports cross-model comparisons, integration with the Azure Developer CLI, and comprehensive documentation tracking.

Capabilities

  • Command-line interface for evaluation management
  • Isolated workspace execution using temporary directories
  • Configurable validation engine with extensible code and regex graders
  • Project scaffolding for new agent skills and evaluation suites
  • Azure Developer CLI (azd) extension support

Best for

Evaluating and benchmarking AI agent skill performance across different models, Scaffolding evaluation suites and test cases for custom agent skills, Running automated compliance and validation checks on agent outputs

Works with

GitHub Copilot
Evoa Score breakdown= Σ (score × weight)
58
Task usefulness20%88 → +17.6

実タスクにどれだけ役立つか(機能の豊富さ・用途の明確さ)。 AIによるcapabilities/use-cases解析

Code quality15%85 → +12.8

実装・指示の品質。 AIによるSKILL.md/README解析

Maintenance15%70 → +10.5

リポジトリがどれだけ活発に保守されているか。 GitHub 最終push日時の新しさ

Documentation12%28 → +3.4

ドキュメントの充実度・分かりやすさ。 README/独自要約の情報量

Security15%80 → +12.0

危険・不審な挙動が無いか。 AIによるセキュリティレビュー

Originality10%80 → +8.0

ありふれたラッパーではない独自性。 AIによる独自性判定

Popularity8%30 → +2.4

コミュニティの採用度。 GitHub Stars/Forks(対数スケール)

Compatibility5%25 → +1.3

対応AIエージェントの広さ。 AIによる対応エージェント判定

Weighted total67.9 / 100

ライセンス不明/制限あり(Red)のSkillは総合スコアに0.85倍の補正を適用します。 ランキングはこのScoreのみで決まり、広告で変わりません。 算出方法の詳細 →

Security considerations

Executes tasks in isolated temporary directories to prevent unintended modifications to source fixtures.

Categories

Summary and analysis are original content generated by AI Skills Rank. The skill's source text is not reproduced here — view it on the linked repository.