Aijo24/mudguard
mudguard is an autonomous, delta-aware architecture analysis skill for Claude Code that sweeps codebases for deep-module and deletion-test opportunities, independently verifying candidates before filing them as vertical-slice issues.
Designed to prevent architectural silt, mudguard operates headlessly to evaluate codebases—either entirely or in specific subsystems—using an inlined methodology derived from modern codebase design principles. It employs a rigorous propose-and-verify split, maintains delta-awareness to avoid re-analyzing unchanged components, and leverages per-area receipts to store exclusion maps and rejection history. The skill runs safely within an isolated git worktree without executing external scripts by default, producing ready-to-grab vertical-slice tickets under a scratch directory.
Sweeping a newly inherited codebase to identify shallow modules and structural bottlenecks, Running delta-aware architectural reviews on active repositories to catch new technical debt, Generating vertical-slice refactoring backlogs for specific subsystems or packages, Independently verifying potential refactoring candidates before turning them into actionable issues
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
実装・指示の品質。 — AIによるSKILL.md/README解析
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
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
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
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
コミュニティの採用度。 — GitHub Stars/Forks（対数スケール）
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
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 →
The skill performs analysis-only operations and writes exclusively to local scratch directories. It executes inside an isolated git worktree on a dedicated branch, avoids pushing changes by default, and downloads or fetches no external data or scripts during standard execution.
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.