Prompt Benchmark

---
name: prompt-benchmark
description: Systematic prompt evaluation framework with MATH, GSM8K, and Game of 24 benchmarks. Use when evaluating prompt effectiveness on standard benchmarks, comparing meta-prompting strategies quantitatively, measuring prompt quality improvements, or validating categorical prompt optimizations against ground truth datasets.
---

# Prompt Benchmark Framework

Systematic evaluation of prompts against standard AI benchmarks.

## Core Benchmarks

| Benchmark | Domain | Metric | Baseline | Meta-Prompt Target |
|-----------|--------|--------|----------|-------------------|
| MATH | Mathematics | Accuracy | ~50% | ≥70% |
| GSM8K | Grade School Math | Accuracy | ~80% | ≥90% |
| Game of 24 | Arithmetic Reasoning | Success Rate | ~30% | ≥90% |
| HumanEval | Code Generation | pass@1 | ~67% | ≥80% |
| MMLU | Knowledge | Accuracy | ~70% | ≥80% |

## Benchmark Implementation

### Game of 24