OKF Semantic-Layer Accuracy Eval
Evaluation report
{{SKILL}}
run
{{RUN_TS}}
tokens via
{{TOKEN_METHOD}}
Summary
Test questions
{{TOTAL_QUESTIONS}}
total evaluated
Overall accuracy
{{OVERALL_PCT}}
{{OVERALL_FRACTION}}
Partially correct
{{PARTIAL_PCT}}
{{PARTIAL_FRACTION}}
Avg tokens / question
{{AVG_TOKENS}}
across all cases
{{CORRECT_N}}
correct · within ±2%
{{PARTIAL_N}}
partial · ±2–5%
{{INCORRECT_N}}
incorrect · beyond ±5%
Case-by-case analysis
{{CASE_BLOCKS}}
Recommended actions
{{RECOMMENDATIONS}}