voice-audit

Detect machine-generated voice markers and authenticity risk in sermon drafts.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/jsschrstrcks1/Romans --skill voice-audit
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: voice-audit
Source: https://github.com/jsschrstrcks1/Romans/tree/main/.claude/skills/voice-audit
Command: npx skills add https://github.com/jsschrstrcks1/Romans --skill voice-audit

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Identifies AI-generated voice markers, authenticity risks, and voice drift in completed sermon and theological manuscript drafts so authors can restore their native pastoral voice without a full rewrite.

Core Features & Use Cases

  • Machine Tell Scan: Grep-style and stylistic checks for structural, transition, word-level, rhythm, substance, image-density, and assumed-familiarity tells.
  • Voice Continuity Check: Compares drafts against an author's sermon corpus for required voice markers and absent patterns, flagging drift with specific locations.
  • Conviction, Cadence & Doctrinal Checks: Tests for pastoral directness, climactic sentence shaping, doctrinal sharpness, and early comfort before conviction.
  • Authenticity Risk Assessment: Aggregates findings into a Low/Medium/High rating and recommends 3–5 minimal, targeted restoration edits rather than wholesale rewrites.
  • Use Case: Post-draft review of a sermon prepared with AI assistance to ensure the final manuscript reads like the verified preacher and retains doctrinal and sermonic integrity.

Quick Start

Run a voice-audit on the attached sermon draft and return flagged machine tells, voice-continuity gaps, conviction and cadence diagnostics, and three minimal restoration edits.

Frequently Asked Questions about voice-audit

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

FAQPage Schema
How do I detect AI-generated text in a sermon draft?

Detect AI-generated text in a sermon draft by scanning for structural, transition, word-level, and rhythm machine tells. This audit identifies authenticity risks and flags specific locations where the manuscript drifts from a natural pastoral voice.

What is a voice continuity check for theological writing?

A voice continuity check for theological writing compares your current draft against an existing sermon corpus to identify required voice markers and absent patterns. It flags voice drift with specific locations to ensure the manuscript maintains your established pastoral style.

How do I restore my pastoral voice after using AI to write a manuscript?

Restore your pastoral voice after using AI by applying targeted, minimal restoration edits rather than a wholesale rewrite. An authenticity audit pinpoints exact machine tells and recommends three to five specific edits to recover doctrinal sharpness and climactic cadence.

Can I audit an AI-assisted sermon for doctrinal sharpness and conviction?

Yes, you can audit an AI-assisted sermon for doctrinal sharpness and conviction by running checks for pastoral directness, climactic sentence shaping, and early comfort before conviction. This detects theological substance tells and verifies sermonic integrity.

Does a machine-tell scan require my previous sermons to work?

A machine-tell scan does not strictly require previous sermons to identify structural and word-level AI markers, but providing an author sermon corpus is necessary for a voice continuity check to compare drafts and flag voice drift accurately.

What are the limitations of using voice audits for manuscript restoration?

The limitation of this voice audit approach is that it avoids wholesale rewrites, focusing instead on providing three to five minimal, targeted restoration edits. It is designed for post-draft reviews and cannot generate new sermonic content or replace missing theological substance.