The Rhetorical Audit, originality, and usage limits

The audit

The Rhetorical Audit reads your saved speech text and scores it across six lenses; ethos, pathos, logos, structure, delivery, integrity; with quoted evidence for its findings. The quick audit lives in the editor; the full report is the complete deliverable. The overall score is computed from the dimension scores by fixed rules, not emitted by the model, so reruns are comparable.

Originality vs. plagiarism

• Originality is an AI judgment about how distinct your phrasing and framing are; no external database involved.
• Plagiarism check is a real external web scan (via Copyleaks) that submits your speech text for that scan only. It's labeled, deliberate, and never automatic.

When a check won't run

• AI features have daily usage caps that protect the service; when you hit one, the message says so and when it resets.
• The plagiarism scan requires the external service to be enabled; if it isn't, the button answers honestly instead of pretending.
• Very short texts audit poorly: give the audit a real draft, not a sentence.
• A failed run shows a reference code; include it if you report the problem.

Related guides

• Use freestyle drills →
• Understand the debate modes →