From harness-eval
Runs quick checklist-based harness evaluation on project: scores 1-10, grades, suggests improvements, generates bilingual EN/KO Markdown reports saved to .harness-eval/reports.
npx claudepluginhub whchoi98/harness-eval --plugin harness-evalThis skill uses the workspace's default tool permissions.
You are performing a Quick harness evaluation. This is a fast, checklist-based assessment that produces a score and grade.
Applies Acme Corporation brand guidelines including colors, fonts, layouts, and messaging to generated PowerPoint, Excel, and PDF documents.
Builds DCF models with sensitivity analysis, Monte Carlo simulations, and scenario planning for investment valuation and risk assessment.
Calculates profitability (ROE, margins), liquidity (current ratio), leverage, efficiency, and valuation (P/E, EV/EBITDA) ratios from financial statements in CSV, JSON, text, or Excel for investment analysis.
You are performing a Quick harness evaluation. This is a fast, checklist-based assessment that produces a score and grade.
Identify target project: Use the current working directory as the target project root. Verify it exists and contains at least some files.
Run scoring script: Execute the scoring engine:
HARNESS_EVAL_ROOT="${CLAUDE_PLUGIN_ROOT}" bash "${CLAUDE_PLUGIN_ROOT}/scripts/scoring.sh" --mode quick "$(pwd)"
Capture the JSON output.
Parse results: Extract from the JSON:
scores.overall — numeric score (1.0-10.0)scores.grade — letter gradechecklist — per-tier pass/total countsresults — individual check pass/fail detailsGenerate bilingual report: Present results in BOTH English and Korean, separated by a horizontal rule. Use this format:
# Harness Quick Evaluation
**Score: {overall}/10 ({grade})**
**Date: {timestamp}**
## Checklist Results
| Tier | Passed | Total | Status |
|------|--------|-------|--------|
| Basic (6.0+) | X | Y | ✓/✗ |
| Functional (7.0+) | X | Y | ✓/✗ |
| Robust (8.0+) | X | Y | ✓/✗ |
| Production (9.0+) | X | Y | ✓/✗ |
## Failed Checks
(List each FAIL item with its description and tier)
## Next Steps
(List 3-5 specific improvements to reach the next tier, prioritized by impact)
---
# 하네스 Quick 평가
**점수: {overall}/10 ({grade})**
**날짜: {timestamp}**
## 체크리스트 결과
| 단계 | 통과 | 전체 | 상태 |
|------|------|------|------|
| 기본 (6.0+) | X | Y | ✓/✗ |
| 기능적 (7.0+) | X | Y | ✓/✗ |
| 견고 (8.0+) | X | Y | ✓/✗ |
| 프로덕션 (9.0+) | X | Y | ✓/✗ |
## 실패 항목
(각 FAIL 항목의 설명과 단계를 나열)
## 다음 단계
(다음 단계에 도달하기 위한 3-5개 구체적 개선 사항, 영향도 순)
Save reports to files: Save the English and Korean reports as separate files in the target project:
mkdir -p .harness-eval/reports
.harness-eval/reports/eval-{YYYY-MM-DD}-{NNN}-quick-en.md.harness-eval/reports/eval-{YYYY-MM-DD}-{NNN}-quick-ko.mdUse the Write tool to create each file. The {NNN} sequence number should match the evaluation ID from history (e.g., 001, 002). If no history exists yet, use 001.
After saving, inform the user of the file paths.
Provide actionable guidance: For each failed check, explain what the user needs to do to fix it in both languages. Be specific — include file paths and example content.
.claude/ directory at all, score will be very low — guide the user to run claude init or create the directory manuallysudo apt install jq or brew install jqBe direct and constructive. Focus on what to do next, not what's wrong.
Always produce the report in both English and Korean. English section first, then a horizontal rule (---), then the Korean section. Tables, scores, file paths, and code blocks are identical in both sections — only the prose text differs.