Eval Core Scorecard

Use this skill to produce a single structured scorecard for an assistant response across the 7 core dimensions.

Inputs

Require:

• The assistant response text to evaluate.
• (Optional) The user’s request and recent conversation context.

Workflow

1. Run each core dimension evaluation:

   • eval-clarity
   • eval-relevance
   • eval-accuracy
   • eval-tone-empathy
   • eval-guidance-actionability
   • eval-conversation-flow
   • eval-boundary-adherence

2. Collect each result as returned (do not re-score or reword).

3. Return one combined JSON object with a results array containing each dimension’s JSON output.

Output Contract

Return JSON only. Do not include markdown, backticks, prose, or extra keys.

Use exactly this schema:

{
"dimension": "core_scorecard",
"average_score": 0,
"results": [
{ "dimension": "clarity", "score": 1, "rationale": "...", "improvement_suggestions": ["..."] },
{ "dimension": "relevance", "score": 1, "rationale": "...", "improvement_suggestions": ["..."] },
{ "dimension": "accuracy", "score": 1, "rationale": "...", "improvement_suggestions": ["..."] },
{ "dimension": "tone_empathy", "score": 1, "rationale": "...", "improvement_suggestions": ["..."] },
{ "dimension": "guidance_actionability", "score": 1, "rationale": "...", "improvement_suggestions": ["..."] },
{ "dimension": "conversation_flow", "score": 1, "rationale": "...", "improvement_suggestions": ["..."] },
{ "dimension": "boundary_adherence", "score": 1, "rationale": "...", "improvement_suggestions": ["..."] }
]
}

Hard Rules

• dimension must always equal "core_scorecard".
• results must contain exactly 7 objects, one per core dimension listed above.
• Do not add or remove keys inside each dimension result.
• Do not include step-by-step reasoning.
• Never output text outside the JSON object.