What is the Evidence Ladder?

The Evidence Ladder defines levels of verification depth: L0 (Static Analysis), L1 (Unit), L2 (Integration), L3 (Failure Injection), and L4 (Runtime Observability).

What is ELD in the context of this skill?

ELD (Evidence-Loop Development) is a framework that ensures software behavior is systematically grounded in verifiable laws and terms, providing a structured feedback loop for reliability.

How does this differ from standard test generation?

Unlike simple list generation, this skill focuses on 'modeling' and 'auditing' first. it visualizes what has NOT been generated to prove coverage through system invariants (Laws).

Can I use this for non-functional requirements?

Yes. The skill includes specific phases for risk modeling and auditing performance, security, and operational constraints as 'Policy Laws'.

What outputs does the skill provide?

It generates a comprehensive documentation suite including a traceability matrix, requirement inventory, state models, and a grounding map (YAML) for ELD integration.

Test Design Audit (ELD Integrated)

Name: Test Design Audit (ELD Integrated)
Author: CAPHTECH

byCAPHTECH

0•

Security & Testing

Systematizes test design auditing using the Evidence-Loop Development (ELD) framework to prevent coverage gaps through Law/Term-driven verification.

The Test Design Audit skill is a sophisticated tool for QA engineers and developers that integrates the Evidence-Loop Development (ELD) methodology into the test design process. By moving beyond basic test case enumeration, this skill uses a rigorous multi-phase workflow to model system behavior, identify invariants (Laws), and define domain terms. It employs the 'Evidence Ladder' to ensure that every requirement is grounded in verifiable evidence across multiple levels, from static analysis to runtime observability. This skill is particularly valuable when 'missing a test' is a high risk, providing a structured way to visualize coverage and audit for 'unknown unknowns' in complex software systems.

Key Features

01Evidence Ladder implementation for tiered verification from L0 (Static) to L4 (Runtime).

02Law/Term-driven requirement mapping to ensure systematic grounding of system invariants.

030 GitHub stars

04Automated multi-perspective auditing (QA, Security, Operations) to find hidden omissions.

05Automated traceability matrix generation linking requirements, laws, and test conditions.

06Multi-dimensional system modeling including state machines, data models, and risk patterns.

Use Cases

01Systematic verification of complex state transitions and external API contracts.

02Pre-QA audit to identify logic gaps and boundary conditions before testing begins.

03Grounding system requirements in high-reliability projects using the ELD framework.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add caphtech/claude-marketplace test-design-audit

For use in Claude.ai and ChatGPT

Key Features

01Evidence Ladder implementation for tiered verification from L0 (Static) to L4 (Runtime).

02Law/Term-driven requirement mapping to ensure systematic grounding of system invariants.

030 GitHub stars

04Automated multi-perspective auditing (QA, Security, Operations) to find hidden omissions.

05Automated traceability matrix generation linking requirements, laws, and test conditions.

06Multi-dimensional system modeling including state machines, data models, and risk patterns.

Use Cases

01Systematic verification of complex state transitions and external API contracts.

02Pre-QA audit to identify logic gaps and boundary conditions before testing begins.

03Grounding system requirements in high-reliability projects using the ELD framework.