Is this skill compatible with standard Claude Code workflows?

Yes, it is designed as a Claude Code skill adapted from ECC intent and translated into OpenClaw-native tool workflows.

What is the primary purpose of the ecc-eval-harness skill?

It is a domain-specific skill used for evaluation, benchmarking, quality scoring, and regression checks to ensure AI agent reliability and performance.

What kind of output does this skill generate?

It provides a standardized report including the task objective, actions taken, verification evidence, risks/limitations, and recommended next steps.

Can this skill be used for regression testing?

Yes, regression checks are a core component of the skill, allowing developers to verify that updates do not degrade agent performance.

How does it handle environments where certain tools are unavailable?

The skill includes OpenClaw constraints that utilize documented fallback scripts in the openclaw-ecc/scripts directory and marks unsupported checks as SKIP.

AI Evaluation Harness

Name: AI Evaluation Harness
Author: oabdelmaksoud

byoabdelmaksoud

0•

安全与测试

Benchmarks and evaluates AI agent performance through structured quality scoring, regression checks, and reliability tuning.

The ecc-eval-harness skill provides a specialized workflow for benchmarking AI agent performance and ensuring consistent quality through rigorous evaluation. Designed for the OpenClaw environment, it translates complex evaluation intent into deterministic tool steps, allowing developers to perform automated regression checks and reliability tuning. It focuses on evidence-based reporting, providing clear verification of outcomes and identifying potential risks or limitations in the agent's behavior during the evaluation process.

主要功能

01Reliability and performance tuning for agent configurations

02Deterministic workflow translation for OpenClaw tools

03Evidence-based verification and regression check management

04Automated quality scoring and performance benchmarking

050 GitHub stars

06Support for fallback scripts and environment-aware execution

使用场景

01Measuring AI agent performance against standard benchmarks after code changes

02Tuning agent reliability and scoring quality for domain-specific tasks

03Automating regression testing to ensure prompt engineering doesn't break existing logic

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add oabdelmaksoud/superclaw ecc-eval-harness

For use in Claude.ai and ChatGPT

主要功能

01Reliability and performance tuning for agent configurations

02Deterministic workflow translation for OpenClaw tools

03Evidence-based verification and regression check management

04Automated quality scoring and performance benchmarking

050 GitHub stars

06Support for fallback scripts and environment-aware execution

使用场景

01Measuring AI agent performance against standard benchmarks after code changes

02Tuning agent reliability and scoring quality for domain-specific tasks

03Automating regression testing to ensure prompt engineering doesn't break existing logic

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add oabdelmaksoud/superclaw ecc-eval-harness

For use in Claude.ai and ChatGPT