More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Creator Voice Provenance Builder

    1

    Build a usable voice guide whose rules trace back to supplied writing samples, declared preferences, and visible conflicts.

    $11

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Creator Voice Provenance Builder

    Creator Voice Provenance Builder

    Example session with this skill installed

    Build a voice profile from these four writing samples. The audience is independent consultants. Prefer conclusion-first writing, avoid exaggerated claims, and preserve the difference between casual working notes and formal client reports.

    • Read your context and instructions
    • Compiled the creator voice provenance

    Demonstration Output — not the buyer's voice

    Status: DEMONSTRATION

    The fictional packet contains four samples and two declared preferences. Observed surface features are recorded separately from interpretation; every usable rule points to its source signals.

    VOICE_PROVENANCE_PROFILE v1.0

    RULE-01
    Instruction: Lead decision-oriented writing with the conclusion or verdict.
    Strength: HARD
    Provenance: SIG-01 → S01, S03
    Conflict: none

    RULE-02
    Instruction: State missing evidence and analytical limits directly.
    Strength: HARD
    Provenance: SIG-02 → S02, S04
    Conflict: none

    RULE-03
    Instruction: Use conversational punctuation in working notes and formal labels in client reports.
    Strength: PREFERRED
    Provenance: SIG-04 → S01; SIG-05 → S03
    Conflict: CONFLICT-01 — register changes by supplied context
    Resolution: SCOPE_SPLIT

    Identity boundary

    Mode: CLEAN_ROOM_ABSTRACTION
    close_imitation: false
    No psychological, demographic, authenticity, or audience-response claim is produced. The profile describes the supplied fictional samples only.

    Compiled voice.md excerpt

    • Lead with the conclusion when a decision is required.
    • Make evidence limits visible.
    • Preserve separate casual-note and formal-report registers.
    • Treat unsupported or contradicted patterns as tests, not hard rules.

    Next action: review CONFLICT-01 and confirm whether the context split should remain in the reusable guide.

    Compiled voice.md

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Build a voice guide with visible evidence

    Most voice guides are collections of adjectives: confident, warm, clear. They rarely show which writing evidence supports a rule, where the creator changes register, or which instructions are preferences rather than observed patterns.

    Creator Voice Provenance Builder turns supplied writing samples into two linked artifacts: a machine-readable provenance profile and a concise voice.md guide. Every rule points to evidence signals, contradictions remain visible, and formal-versus-casual variation can be scoped instead of averaged away.

    What it delivers

    • Stable sample, signal, rule, preference, and conflict IDs
    • Observed surface measurements separated from interpretation
    • HARD, PREFERRED, and EXPERIMENTAL writing rules with provenance links
    • Explicit preference-conflict resolutions
    • A clean-room boundary for named creator references
    • A reusable voice.md plus optional version-difference record

    Boundaries

    It does not write a campaign, clean one draft, clone a named living creator, infer psychological or demographic identity, or promise audience response. One or two samples produce a provisional profile. If no sample is supplied, the Skill runs a labelled fictional Demonstration Output.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Prompt & Skill Engineering.

    A profile describes supplied writing, not the full person. One or two samples produce a PROVISIONAL profile. Frequency does not prove identity, quality, intent, authenticity, or audience impact. Named living creators are handled through clean-room structural abstraction; close imitation and identity substitution are refused. Validation proves contract and provenance links, not cross-model voice fidelity.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 2 months ago

    Permissions

    Terminal / Shell
    Read Files
    Write Files

    File Scopes

    *.txt
    *.json
    *.md

    Uses local Python scripts to read user-selected writing samples or prior profiles and write the requested JSON profile, voice.md guide, optional observation ledger, and optional version diff. No browser, network, external host, or environment-variable access is required.

    Listed2 months ago

    What's inside

    Frequently Asked Questions

    Unlike standard guides that use vague adjectives, this skill creates a 'provenance profile' where every stylistic rule is linked to specific evidence in your writing samples, ensuring your AI output is grounded in actual patterns rather than guesses.

    You will receive the machine-readable provenance profile, a formatted `voice.md` guide for immediate use, and a conflict-resolution log that identifies where your writing style may contradict your stated preferences.

    The skill requires at least one writing sample to function; while it can run a demonstration using fictional data, a minimum of three diverse samples is recommended to generate a stable, accurate profile.

    The output is delivered in standard Markdown (.md) and JSON-compatible formats, making it compatible with any LLM, custom GPT, or AI writing assistant that supports system instructions or style injections.

    Yes, the skill creates a 'version-difference record' allowing you to update your profile as your writing style evolves while maintaining a clear history of how your voice rules have changed over time.

    New in Prompt & Skill Engineering