skills.rest
AesopScott
AesopScottCOMMUNITY
Explore all skills by this author →
GitHub

verification-checks

Define failure modes, test cases, and observability metrics for AI system harnesses.

Updated May 11, 2026
Search Tags:#validation#verification#compliance#safety#trust#harness
One-click install
npx skills add https://github.com/AesopScott/mojo --skill verification-checks
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: verification-checks
Source: https://github.com/AesopScott/mojo/tree/main/harnesses/skills/verification-checks
Command: npx skills add https://github.com/AesopScott/mojo --skill verification-checks

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill addresses the risk of unverified AI outputs by providing a structured framework for implementing scripted checks and evidence schemas, ensuring that only validated responses influence downstream operations.

Core Features & Use Cases

  • Capability Control: Define strict boundaries for what the AI is permitted to change or execute.
  • Evidence-Based Validation: Require passing tests and audit logs before confirming provider, route, or response integrity.
  • Use Case: When modifying a critical system harness, use this Skill to define a positive test and a boundary test that prevents the AI from accidentally modifying production secrets or bypassing approval gates.

Quick Start

Use the verification-checks skill to define a new evidence schema and validation test for the current harness boundary.

Frequently Asked Questions about verification-checks

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

FAQPage Schema
How do I implement validation and compliance checks for an AI system harness?▼

To implement validation checks for an AI system harness, you must define a rigorous verification framework that establishes capability control boundaries, requires evidence-based test passing, and generates audit-ready observability metrics.

What is evidence-based validation for AI-driven workflows?▼

Evidence-based validation is a gating mechanism that requires passing specific positive and boundary test cases, alongside audit logs, before confirming provider or response integrity to ensure only validated outputs proceed downstream.

How do I prevent AI from accidentally modifying production secrets in automated workflows?▼

You prevent unauthorized modifications by defining strict capability control boundaries and specific failure modes, applying boundary tests that restrict the AI from bypassing approval gates or accessing production secrets.

Do I need predefined failure modes to establish trust and safety boundaries?▼

Yes, defining specific failure modes is a mandatory requirement for establishing operational safety boundaries, as it dictates the positive and boundary test cases needed to validate response integrity.

How do I set up observability metrics for response validation?▼

You set up observability metrics by defining a new evidence schema and validation tests for your current harness boundary, ensuring the framework generates audit-ready logs for every capability-controlled workflow.

Related Skills

Explore More Software Engineering Skills→
AesopScottAesopScott

process-hooks

Define artifact schemas and post-process hooks for AI response handling.

Community
Advanced
ybbms777ybbms777
4

safe-exec

Detect and mitigate dangerous shell commands with BashAST-based risk scoring.

Community
Advanced
aibangjuxinaibangjuxin
2

api-tester

Test RESTful APIs with manual and automated scripts.

Community
Intermediate
chiubakachiubaka
1

runtime-boundary-validation-zod

Validate untrusted JSON data at runtime using Zod schemas.

Community
Advanced