Analyzes and restores a writer's stylistic fingerprint in AI-assisted writing workflows.
npx skills add https://github.com/TylerGarlick/abraxas --skill ethos
Please help me install this Agent Skill. Skill: ethos Source: https://github.com/TylerGarlick/abraxas/tree/main/skills/ethos Command: npx skills add https://github.com/TylerGarlick/abraxas --skill ethos
This Skill combats "voice drift" in AI-assisted writing by capturing, monitoring, and restoring a writer's unique stylistic fingerprint.
/ethos register to capture your original voice, then /ethos check on new drafts to see the drift score and get suggestions for restoration using /ethos restore.Register your current writing style by providing a text sample using the command '/ethos register'.
High-intent search queries and answers about installing and using this skill.
To prevent voice drift, register a baseline text sample of your writing style for analysis across sentence structure, vocabulary, rhythm, and tone. This creates a stylistic fingerprint to compare future AI-assisted drafts against and detect style deviations.
A writing style fingerprint is a baseline profile of your text captured across vocabulary, sentence structure, rhythm, and tone dimensions. It is analyzed by comparing new drafts against this registered baseline to identify deviations in your authentic voice.
Run a drift detection check on your new drafts to compare them against your registered stylistic fingerprint. This generates a clear drift score and provides actionable recommendations to correct any detected voice drift.
You can restore your writing voice by applying algorithmic or guided rewriting to bring drifted text back in line. This correction process targets specific style deviations identified during your baseline fingerprint comparison.
Voice drift detection integrates with Janus to selectively analyze Sol factual content while excluding Nox creative or dream content. This ensures style monitoring applies only to your appropriate factual writing workflows.
Track voice health over time by reviewing your audit history, which logs drift scores from your baseline fingerprint. This history identifies specific writing sessions or edits that caused stylistic deviations for targeted restoration.
cdeistopenedExtract a writer's voice from samples into a reusable voice-dna.md fingerprint.
zeronx798Identifies and removes AI-generated writing patterns to restore natural prose.
TylerGarlickLabel AI claims with confidence levels and separate factual from symbolic output.
frankxaiIdentify and refine an authorial voice through rhythm, vocabulary, and perspective analysis.