Single tool 'execute_skill_action' has critical definition gaps. Description is vague (79 chars, below 100-char minimum for clarity) and does not explain WHAT the tool does, WHEN to use it, or WHAT it returns. Input schema exists but parameter 'query_payload' lacks actionable description, it is a free-form string with no format guidance, constraints, or examples. No output schema is documented; the tool returns a generic JSON object with 'status' and 'message' fields, but the LLM cannot predict structure or downstream chaining. No error handling guidance. Tool name 'execute_skill_action' is generic and does not follow verb_noun pattern (e.g., 'evaluate_llm_output', 'score_agent_response'). The server claims to be an 'Automated LLM Judge Feedback Evaluator' but the tool name and description do not reflect this domain purpose.

• Rename tool to 'evaluate_llm_output' or 'score_agent_response' to match verb_noun pattern and clarify domain purpose.
• Expand tool description to 120-180 chars: 'Evaluates an LLM agent's output against a rubric of criteria (e.g., Coherence, Accuracy, Conciseness). Returns per-criterion scores (0-5), aggregate score, and pass/fail status. Use this to assess output quality before returning to user.'
• Replace 'query_payload' parameter with explicit fields: 'input_prompt' (string, the original task), 'agent_output' (string, the LLM's response), 'rubric_criteria' (array of strings, evaluation dimensions). Add descriptions for each.
• Document output schema: 'Returns object with evaluation_id (string), criteria_breakdown (object mapping criterion name to {score, max, feedback}), aggregate_score (number 0-5), pass_threshold_met (boolean), judge_eval_dashboard_url (string).'
• Add error handling: return 'Invalid rubric_criteria: must be non-empty array' if criteria is empty, 'Missing required field: input_prompt' if prompt is absent, etc. Include recovery guidance.