linshenkx/prompt-optimizer · error · EvaluationParseError

Evaluation result is missing score for "${fieldName}".

Error message

Evaluation result is missing score for "${fieldName}".

What it means

EvaluationParseError thrown by the inner extractScore helper when a specific dimension/sub-score (fieldName) inside the evaluation result is undefined or null. It fires while extracting individual criterion scores, naming the offending field in the message. The top-level score check has already passed by this point.

Source

Thrown at packages/core/src/services/evaluation/service.ts:3123

   * 标准化评估响应（统一结构）
   */
  private normalizeEvaluationResponse(
    data: any,
    type: EvaluationType,
    metadata?: EvaluationResponse['metadata']
  ): EvaluationResponse {
    if (!data || typeof data !== 'object') {
      throw new EvaluationParseError('Evaluation result is not a valid object.');
    }

    if (data.score === undefined || data.score === null) {
      throw new EvaluationParseError('Evaluation result is missing the "score" field.');
    }

    // 提取分数（0-100，整数）
    const extractScore = (value: any, fieldName: string): number => {
      if (value === undefined || value === null) {
        throw new EvaluationParseError(`Evaluation result is missing score for "${fieldName}".`);
      }
      const num = typeof value === 'number' ? value : parseInt(String(value));
      if (isNaN(num)) {
        throw new EvaluationParseError(`Invalid numeric score for "${fieldName}": ${value}`);
      }
      return Math.max(0, Math.min(100, num));
    };

    const tryExtractScore = (value: any, fieldName: string): number | null => {
      try {
        return extractScore(value, fieldName);
      } catch {
        return null;
      }
    };

    const toDimension = (key: string, label: string, scoreValue: any): EvaluationDimension | null => {
      const score = tryExtractScore(scoreValue, `dimension.${key}`);

View on GitHub (pinned to 3e677b1d9f)

Solutions

1. Inspect the message for the missing fieldName and check the raw model output for that key
2. Require every dimension in the judge prompt explicitly ('you MUST provide a numeric score for each of: ...')
3. Increase max_tokens so trailing dimensions aren't truncated

const dims = ['accuracy','relevance','clarity']; // assert before/at prompt time
const missing = dims.filter(d => parsed?.dimensions?.[d] == null);

const hasAllDimensionScores = (v: any, dims: string[]): boolean => dims.every(d => v?.dimensions?.[d] != null);

try { ... } catch (e) { if (e instanceof EvaluationParseError && /missing score for/.test(e.message)) { /* re-prompt with mandatory dimensions, retry */ } else throw e; }

When it happens

Trigger: A multi-dimension evaluation result where one dimension (e.g. accuracy, relevance) is absent or null in the JSON, e.g. {"score": 80, "dimensions": {"accuracy": null }}.

Common situations: Judge model skipping dimensions it deems not applicable; prompt listing dimensions the model doesn't echo back; partial JSON truncation dropping trailing fields.

Related errors

• Evaluation result is missing the "score" field.
• Invalid numeric score for "${fieldName}": ${value}
• Failed to parse evaluation result: no valid score JSON or re
• Evaluation result is not a valid object.
• Evaluation result is missing a valid overall score.

AI-assisted analysis of linshenkx/prompt-optimizer@3e677b1d9f (2026-08-27). Data as JSON: /api/errors/7efc3bae53e1603f. Report an issue: GitHub.