メニュー
Display a terminal scorecard of benchmark results showing pass rates, scores by difficulty, and per-test breakdowns. Use when the user asks about benchmark results, scores, or wants to see how their SDK performed.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/PSPDFKit-labs/agentic-usability --skill reportコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
SKILL.md を表示中
Initialize a new agentic-usability benchmark pipeline project. Use when setting up a new SDK benchmark, creating a config.json, or starting a new evaluation project.
Launch an interactive shell inside a microsandbox for debugging. Supports bare mode, executor setup, or judge setup with optional test case scaffolding.
Run the full evaluation pipeline (execute, judge, report) for an SDK usability benchmark. Use when running a complete benchmark end-to-end, resuming an interrupted pipeline, or checking pipeline status.
SOC 職業分類に基づく
| name | report |
| description | Display a terminal scorecard of benchmark results showing pass rates, scores by difficulty, and per-test breakdowns. Use when the user asks about benchmark results, scores, or wants to see how their SDK performed. |
| argument-hint | [project-directory] [--json] [--run runId] |
| allowed-tools | Bash(agentic-usability *) Read Glob |
Display the benchmark scorecard for the pipeline.
agentic-usability report -p $ARGUMENTS
--json: Output raw structured JSON instead of the colored table--run <runId>: Show results for a specific run (default: latest)results/<runId>/
report.json # Aggregate scorecard
<targetName>/<testId>/
judge.json # Per-test judge scores
generated-solution.json # Agent's solution
agent-notes.md # Agent's working notes
| Dimension | Range | What it measures |
|---|---|---|
apiDiscovery | 0-100 | Found correct SDK endpoints/methods? |
callCorrectness | 0-100 | API calls constructed correctly? |
completeness | 0-100 | All requirements handled? |
functionalCorrectness | 0-100 | Code runs and produces correct output? |
overallVerdict | boolean | Solution works? (pass/fail) |
The report aggregates these across all test cases and breaks them down by difficulty (easy/medium/hard).
Runs are stored as subdirectories in results/ containing run.json:
{ "id": "run-2026-04-25T10-30-00-000Z", "createdAt": "...", "targets": [...], "testCount": 15, "label": "..." }
To list all runs, look for results/*/run.json files.
Present the results to the user. If they want deeper analysis, suggest using the insights skill.
For detailed file inventory, see pipeline-guide.md.