hardening-pass

Detect and fix quality issues in task templates and task instances before benchmarking.

35|5|Updated Apr 11, 2026
One-click install
npx skills add https://github.com/TheodoreGalanos/aec-bench --skill hardening-pass
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hardening-pass
Source: https://github.com/TheodoreGalanos/aec-bench/tree/main/src/aec_bench/init/skill_data/hardening-pass
Command: npx skills add https://github.com/TheodoreGalanos/aec-bench --skill hardening-pass

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Review and harden a task template or task instance for correctness, robustness, and benchmark readiness, ensuring validity before use in experiments.

Core Features & Use Cases

  • Guardrails for template and task quality to prevent invalid benchmarks.
  • Detect and address issues in metadata, instructions, and environment configurations.
  • Use during template creation, template updates, or pre-benchmark reviews to generate a formal hardening report.

Quick Start

Provide a path to a template or task directory to run a hardening pass and receive a structured report.

Frequently Asked Questions about hardening-pass

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

FAQPage Schema
How do I validate task templates before running benchmarks?

To validate task templates before benchmarking, run a hardening pass to detect quality issues and enforce checks for metadata, formula correctness, parameter realism, and instruction clarity.

What is a hardening pass for benchmark tasks?

A hardening pass is a quality gate that reviews new or updated task templates and instances to ensure validity, robustness, and reproducibility across scenarios before they are used in experiments.

How do I check task instruction clarity and parameter realism?

You can check instruction clarity and parameter realism by providing a task directory path to a template validation process, which generates a structured report highlighting these specific quality issues.

Can I review environment and verifier integration for task instances?

Yes, template hardening enforces checks for proper environment and verifier integration, ensuring your task instances are correctly configured and robust before entering the benchmarking phase.

What is the best way to prevent invalid benchmarks from poorly configured templates?

The best way to prevent invalid benchmarks is to apply a quality gate that detects and addresses issues in metadata, instructions, and environment configurations before executing experiments.

Do I need any dependencies to run a task template quality review?

No external dependencies are required to run a task template quality review, as the hardening process operates directly on the provided template or task directory path to generate its report.