Loading ToolBench
Server has 4 tools with basic schemas and descriptions in Japanese. All tools have input schemas with type definitions and descriptions present. However, descriptions are minimal (10-100 chars, well below the 194-char baseline), lack actionable context for LLM selection, and provide no guidance on when to use each tool vs. alternatives. Parameter descriptions are present but terse. No output schemas documented. Error handling is absent, tools return raw text without recovery guidance. Tool naming follows verb_noun pattern (dice_roll, github_*) but lacks clarity on distinctions between similar GitHub tools (daily vs. comparison vs. repo stats). No tool annotations (readOnlyHint, etc.) despite all being READ_ONLY. Composition is weak: three GitHub tools operate on overlapping concerns without clear separation of responsibilities.
サイコロを振った結果を返す
GitHubユーザーの今日の活動を過去7日平均と比較
GitHubユーザーの今日のコミット有無と統計を返す
GitHubユーザーの対象レポジトリでの直近のコミット統計（追加・削除行数）を返す
Descriptions are too brief and lack LLM-optimized context. 'サイコロを振った結果を返す' (10 chars) and 'GitHubユーザーの今日のコミット有無と統計を返す' (25 chars) fall far below the 194-char baseline. No guidance on when to call each GitHub tool instead of alternatives.
Three GitHub tools (github_repo_commit_stats, github_daily_commit_stats, github_commit_comparison) operate on overlapping concerns without clear separation. LLM cannot distinguish when to use 'daily' vs 'comparison' vs 'repo stats'. Violates single-responsibility principle.
No output schemas documented. Tools return unstructured text (e.g., 'content: [{ type: "text", text: ... }]'). LLM cannot plan downstream calls or extract structured data. Violates response-shaper pattern.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 51 | 2026-07-28+ | v2 |
No error handling or recovery guidance. Tools return raw text on failure with no actionable next steps. Missing error classification (retryable vs user-fixable vs fatal) and recovery hints.
Tool annotations missing. All tools are READ_ONLY but lack readOnlyHint annotation. Prevents agents from understanding safety properties and optimizing execution strategy.