What is confidence grading in this context?

Confidence grading labels findings as High (multiple points of evidence), Medium (strong evidence with slight ambiguity), or Low (weak or incomplete evidence) to assist in prioritization.

What is the Audit Fleet Evidence Policy?

It is a specialized Claude Code skill designed to enforce a strict structure and quality standard for reporting audit findings, ensuring every issue is verifiable and actionable.

Why would Claude reject a finding when using this skill?

The skill includes rejection criteria for any findings that miss required keys or use non-standard severity levels, maintaining the integrity of the audit data.

How does the skill handle evidence collection?

It requires every finding to include a file path, line number, and a short quote or command output to ensure findings are traceable to the source.

What severity levels are supported by this skill?

The skill strictly enforces a severity enum containing only three values: critical, warning, and info.

Audit Fleet Evidence Policy

Name: Audit Fleet Evidence Policy
Author: nsalvacao

bynsalvacao

0•

安全与测试

Enforces rigorous quality control and standardized evidence reporting for software audits and system findings.

The Audit Fleet Evidence Policy skill provides a governance framework for Claude Code to ensure all audit findings are verifiable, traceable, and decision-ready. It mandates a strict schema for every entry, requiring specific metadata such as severity enums, file-line evidence, impact assessments, and confidence grading. By implementing automated rejection criteria for incomplete data, this skill transforms raw AI observations into professional, production-grade reports that meet the high standards of security, reliability, and delivery audits.

主要功能

01Strict schema enforcement for finding keys including impact, owner, and acceptance criteria

020 GitHub stars

03Evidence-based verification requiring file paths, line numbers, and code snippets

04Automated report structuring from executive summaries to actionable next steps

05Standardized severity mapping restricted to critical, warning, and info levels

06Three-tier confidence grading system to quantify finding reliability

使用场景

01Enforcing consistent documentation standards for automated QA and peer review processes

02Standardizing security and reliability findings across large-scale repository audits

03Generating decision-ready technical debt reports with explicit ownership and impact

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add nsalvacao/nsalvacao-claude-code-plugins audit-fleet-evidence-policy

For use in Claude.ai and ChatGPT

主要功能

01Strict schema enforcement for finding keys including impact, owner, and acceptance criteria

020 GitHub stars

03Evidence-based verification requiring file paths, line numbers, and code snippets

04Automated report structuring from executive summaries to actionable next steps

05Standardized severity mapping restricted to critical, warning, and info levels

06Three-tier confidence grading system to quantify finding reliability

使用场景

01Enforcing consistent documentation standards for automated QA and peer review processes

02Standardizing security and reliability findings across large-scale repository audits

03Generating decision-ready technical debt reports with explicit ownership and impact

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add nsalvacao/nsalvacao-claude-code-plugins audit-fleet-evidence-policy

For use in Claude.ai and ChatGPT