Prompt Evaluation Form

Fill out the following evaluation form for prompt validation:
Input format:
[
"evaluator": "Your name/ID",
"promptTitle": "Title of the prompt being evaluated",
"promptID": link,
"scores": [
"clarity": 1-5,
"relevance": 1-5,
"structure": 1-5,
"taskAppropriateness": 1-5,
"efficiency": 1-5
],
"strengths": [
"Strength 1 or .",
"Strength 2 or ."
],
"weaknesses": [
"Weakness 1 or .",
"Weakness 2 or ."
],
"improvements": [
"Improvement 1 or .",
"Improvement 2 or ."
],
"recommendation": "adopt|refine|reject",
"comments": [
"Comment 1 or .",
"Comment 2 or ."
]
]

Output will be formatted as:

### **Human Evaluation Form for Prompt Validation**
#### **Section 1: Basic Information**
1. **Evaluator Name/ID:** <<evaluator>>
2. **Date of Evaluation:** <<CURRENT_DATE>>
3. **Prompt Title:** <<promptTitle>>
---
#### **Section 2: Quantitative Evaluation (Rubric-Based)**
Evaluate the prompt based on the following criteria using the provided scale:
| Criterion | 1 (Very Poor) | 2 (Poor) | 4 (Good) | 5 (Excellent) | Score |
|-------------------------|---------------------|----------------------|----------------------|----------------------|-------|
| **Clarity** | Unclear instructions; difficult to understand | Mostly clear but some confusion | Generally clear; minor ambiguities | Perfectly clear with no room for misinterpretation | <<scores.clarity>> |
| **Relevance** | Completely irrelevant to the task | Some relevance but needs significant revision | Relevant but minor improvements required | Highly relevant and perfectly aligned | <<scores.relevance>> |
| **Structure** | No structure; very disorganized | Basic structure but lacks coherence | Well-structured but needs refinement | Perfectly logical and systematic | <<scores.structure>> |
| **Task Appropriateness**| Misaligned with task requirements | Partially aligned but significant mismatch | Generally matches requirements with minor misalignments | Perfectly aligns with task requirements | <<scores.taskAppropriateness>> |
| **Efficiency** | Inefficient; wastes tokens/time | Somewhat inefficient but performs basic functions | Generally efficient with minor improvements needed | Highly efficient and optimized | <<scores.efficiency>> |
**Total Score:** <<CALCULATED_TOTAL>> / 25
**Percentage:** <<CALCULATED_PERCENTAGE>>%
---
#### **Section 3: Qualitative Evaluation**
1. **Key Strengths:**
- <<if strengths[0]==".">>No specific strength noted<<else>><<strengths[0]>><<endif>>
- <<if strengths[1]==".">>No specific strength noted<<else>><<strengths[1]>><<endif>>
2. **Key Weaknesses:**
- <<if weaknesses[0]==".">>No specific weakness noted<<else>><<weaknesses[0]>><<endif>>
- <<if weaknesses[1]==".">>No specific weakness noted<<else>><<weaknesses[1]>><<endif>>
3. **Suggested Improvements:**
- <<if improvements[0]==".">>No specific improvement suggested<<else>><<improvements[0]>><<endif>>
- <<if improvements[1]==".">>No specific improvement suggested<<else>><<improvements[1]>><<endif>>
---
#### **Final Recommendation**
1. **Overall Assessment:**
- <<if recommendation=="adopt">>☑<<else>>☐<<endif>> Adopt with minor adjustments
- <<if recommendation=="refine">>☑<<else>>☐<<endif>> Refine and retest
- <<if recommendation=="reject">>☑<<else>>☐<<endif>> Reject; requires significant revision
2. **Additional Comments:**
- <<if comments[0]==".">>No additional comment<<else>><<comments[0]>><<endif>>
- <<if comments[1]==".">>No additional comment<<else>><<comments[1]>><<endif>>


Example usage:
[
"evaluator": "John Doe",
"promptTitle": "Image Description Generator",
"scores": [
"clarity": 4,
"relevance": 5,
"structure": 4,
"taskAppropriateness": 5,
"efficiency": 4
],
"strengths": [
"Clear and concise instructions for image description",
"."
],
"weaknesses": [
"Could benefit from more examples",
"."
],
"improvements": [
"Add 2-3 diverse examples of good descriptions",
"."
],
"recommendation": "adopt",
"comments": [
"Overall strong prompt with minor room for improvement",
"."
]
]

Note:
1. The following fields will be automatically generated:
- <<CURRENT_DATE>>: Will be filled with the current date in YYYY-MM-DD format
- <<CALCULATED_TOTAL>>: Sum of all scores
- <<CALCULATED_PERCENTAGE>>: (Total score / 25) * 100
2. For short-answer fields, you can input "." to indicate no response, which will be replaced with an appropriate default message