harness-score-analyzer

Name: harness-score-analyzer
Availability: InStock
Author: bianhaifeng789-hue

Analyze QA score histories to categorize trends and recommend strategies.

Updated Apr 14, 2026

Search Tags:#qa #strategy #analyze #pivot #refine #trend #scores

Ready to run

npx skills add https://github.com/bianhaifeng789-hue/openclaw-config --skill harness-score-analyzer

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: harness-score-analyzer
Source: https://github.com/bianhaifeng789-hue/openclaw-config/tree/main/skills/harness-score-analyzer
Command: npx skills add https://github.com/bianhaifeng789-hue/openclaw-config --skill harness-score-analyzer

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Analyzes QA score histories to identify trends (IMPROVING, STAGNANT, DECLINING) and informs actionable strategies to improve performance.

Core Features & Use Cases

Trend detection: categorize score history into IMPROVING, STAGNANT, or DECLINING.
Strategy recommendations: suggest REFINE or PIVOT based on delta thresholds.
Lightweight statistics: compute mean, max, min, and basic forecast for next iteration.
Use Case: apply in multi-round QA cycles to guide subsequent actions and experiments.

Quick Start

Provide a sample score history like 7.5, 8.0, 7.8 to see the analyzer produce a trend and recommended strategy.

Frequently Asked Questions about harness-score-analyzer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I analyze QA score trends to guide my next iterative workflow steps?▼

Analyzing QA score trends involves evaluating multi-round score histories to categorize performance as IMPROVING, STAGNANT, or DECLINING, which then guides subsequent actions. This process computes core statistics to offer actionable predictive guidance for your next steps.

What is the best way to decide whether to refine or pivot my strategy based on stagnant scores?▼

Deciding whether to refine or pivot your strategy depends on evaluating delta thresholds within your score history. When scores categorize as STAGNANT, analyzing these historical deltas recommends either a REFINE or PIVOT strategy to improve performance.

Can I compute lightweight statistics and forecasts from a basic score history sample?▼

Yes, you can compute lightweight statistics and a basic forecast from a basic score history sample. Providing a simple series like 7.5, 8.0, 7.8 calculates the mean, max, min, and a predictive forecast for the next iteration.

How do I detect if my multi-round QA cycles are improving or declining over time?▼

Detecting if multi-round QA cycles are improving or declining requires categorizing the score history trajectory. The analyzer evaluates the sequence of scores to determine if the trend is IMPROVING, STAGNANT, or DECLINING over time.

Does this analyzer work without external dependencies for multi-round QA cycles?▼

Yes, this analyzer works without external dependencies for multi-round QA cycles. It uses a frontmatter configuration to compute core statistics, categorize trends, and offer simple predictive guidance directly based on the provided score data.