voice-drift-detector

Detect ghostwriting voice drift by comparing recent posts against client reference posts to calculate a cosine similarity score. This audit measures signature-phrase usage, sentence-structure, vocabulary, decision-frame, and story-shape consistency to identify deviations.

Cosine similarity analysis for content quality audits measures the mathematical distance between recent ghostwritten content and a client reference set. It scores voice consistency across specific axes like vocabulary adherence, signature-phrase usage, and story-shape consistency.

Audit client voice consistency weekly by running a per-client scope comparison of recent posts against reference posts. The process validates inputs against client profiles, computes cosine similarity across five axes, and emits a concise brief plus a machine-readable audit record.

Yes, voice drift detection works for multiple active clients by applying the audit to a per-client scope. It validates each client's inputs against their specific profiles and reference sets to produce individualized similarity scores and machine-readable audit records.

The best way to measure signature-phrase usage is through an axis-based audit comparing recent ghostwritten posts against client reference posts. This detects deviations in the client's specific vocabulary and phrase patterns using cosine similarity scoring.

Before running a voice-drift audit, you need a client profile and a set of reference posts written by the client. These inputs are validated against each other to ensure accurate cosine similarity computation and axis analysis for recent posts.