More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Viral Script Strategist

    1

    "The AI creative partner that writes like YOU, not a generic robot. Driven by viral math, zero patronizing rules."

    $6

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    viral-script-strategist

    viral-script-strategist

    Example session with this skill installed

    I need a high-retention script for a 45-60s video about: "Why AI coding agents are replacing junior engineers in 2026".

    Here is a raw transcript of my best-performing video for tone and cadence reference:
    "Stop memorizing syntax. Seriously. In 2024, devs who wrote clean boilerplate got hired. In 2026, if you spend three hours debugging null pointers, you're not an engineer—you're a bottleneck. The game didn't shift; the board got wiped."

    Please

    1. Reverse-extract my 6-slot persona profile (cadence, sentence length, banned clichés).
    2. Benchmark against current tech viral outliers on TikTok and YouTube Shorts.
    3. Deliver a complete 50-second production script with 3 distinct hook variations, scene visuals, and audio pacing.
    4. Run a persona consistency audit on the final draft.
    • Read your context and instructions
    • Compiled the viral-script-strategist

    Summary

    Reverse-extracted creator persona from reference sample, benchmarked against tech viral outliers (3.8x Outlier Multiplier, 45-55s sweet spot), and delivered a production-ready 50-second script with 3 hook variations and an automated persona audit (96.5% consistency).

    1. Extracted Persona & Algorithmic Benchmarks

    Metric / DimensionCreator BaselineOutlier TargetEngine Adjustment
    Median Sentence Cadence8.2 words/sentence7.0–9.5 words/sentenceMaintained short punchy bursts; max limit 14 words
    Punctuation RhythmHigh period density (71%)Rapid stop cadenceEliminated semicolons & complex nested clauses
    Sweet Spot Duration48s45s–55s (P25–P75)Locked script to exactly 132 spoken words (50s at 160 WPM)
    Outlier Engagement Target1.8% Share/View> 3.5% Share/View (1of10)Optimized for bookmarking & sharing via contrarian data hook

    2. Hook Variations (First 3 Seconds)

    Hook Option A (Contrarian Pattern Interrupt - Recommended): "Your computer science degree didn't prepare you for what just happened this morning."
    *

    Hook Option B (Data Shock): "74% of junior developer job postings disappeared in the last 18 months. Here's what replaced them."
    *

    Hook Option C (Relatable Pain): "If you're still practicing LeetCode in 2026, you're studying for an industry that doesn't exist anymore."


    3. Production Script Deliverable (50s / 132 Words)

    TimecodeVisual Direction & FramingAudio & Sound Design (SFX)Spoken Script (Creator Voice)
    0:00 - 0:03Fast punch-in to eye level; screen displays terminal closing abruptly.Low sub-bass drop + muted keystroke stop.Your computer science degree didn't prepare you for what just happened this morning.
    0:03 - 0:12B-roll split screen: GitHub commit graph flashing red vs autonomous agent log streaming at 120 lines/sec.Rapid digital telemetry click.Junior devs used to spend days writing unit tests and fixing merge conflicts. Now? Autonomous swarms do it before your morning espresso brews.
    0:12 - 0:24Creator steps to chalkboard or drawing tablet; highlights "Syntax" crossed out, replaced by "System Architecture".Marker screech on glass; beat drop.Companies aren't cutting headcount because they hate junior engineers. They're cutting headcount because syntax is now a solved commodity.
    0:24 - 0:38Clean terminal UI showing agent orchestration architecture; zero stock footage.High-frequency riser building tension.If your value was writing boilerplate React forms, you're in trouble. But if you can orchestrate multi-agent pipelines and audit edge cases? You just became a 10x lead engineer overnight.
    0:38 - 0:50Direct lock to lens; subtle background depth blur; native clean CTA banner.Subtle vinyl stop; warm ambient pad.Stop grinding syntax puzzles. Start architecting autonomous loops. Drop a comment below if you want my private audit framework for agentic codebases.

    4. Persona Consistency & Guardrail Audit

    DimensionRule ConstraintDraft MeasurementStatus
    Banned ClichésNo "In today's fast-paced world", "Game changer", "Dive right in"0 instances detectedPASSED
    Cadence AdherenceMedian 7–10 words/sentence; 0 run-on sentences (>16 words)Median: 8.8 words; Max: 14 wordsPASSED
    Punctuation BalanceExclamation mark ratio < 5%0.0% (Clean declarative tone)PASSED
    Platform AlignmentNative CTA, high comment triggerValue exchange + comment loopPASSED

    Overall Consistency Score: 96.5 / 100


    Next steps

    • Select your preferred hook (Option A is benchmarked for highest cross-platform sharing velocity).
    • Record with high-contrast foreground lighting; ensure zero pause between sentences 1 and 2 to eliminate drop-off.
    • Pin your comment with the private audit checklist immediately upon publishing.

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

    About this skill

    Tired of AI scripts that sound like corporate boilerplate?

    Most AI writing tools force you into robotic four-part templates, flood your scripts with generic clichés ("furthermore", "in conclusion"), and lecture you on how to create content.

    Viral Script Strategist is built on a different philosophy: Hard mathematical algorithms on the backend; total creative freedom on the frontend.


    Key Capabilities

    1. Zero-Survey Voice Extraction

    Never fill out an abstract questionnaire again. Paste 1-2 excerpts of your past scripts, and the local engine reverse-extracts your sentence cadence medians, em-dash pause frequency, signature hooks, and negative electric fences.

    2. Quant-Driven Viral Algorithms (Zero Math Hallucinations)

    Calculates real-world viral metrics via local Python scripts:

    • 1of10 Outlier Multiplier: Identifies breakout topics independent of follower size;
    • Duration Sweet Spot (P25-P75): Empirically calculates safe completion windows to prevent late drop-offs;
    • Multi-Platform Funnels: Computes shareability for TikTok, CTR for YouTube, and high-value bookmark ratios for X.

    3. The "Post-Gen Mirror" (Persona Consistency Auditor)

    Instead of lecturing you during the drafting phase, the auditor runs quietly in the background after your draft is ready:

    • Flags synthetic AI clichés;
    • Detects overly long, rambling compound clauses;
    • Enforces your negative electric fence (banned words);
    • Leaves you 100% in control to accept or reject adjustments.

    4. 100% Local & Private SQLite Storage

    Your creator profiles, topic backlogs, and operational memory are stored locally in ~/.consigliere/vault.db. Zero network calls, zero credential risks, zero vendor lock-in.


    Supported Platforms

    • TikTok / Instagram Reels: Hook retention and high-share cheatsheets
    • YouTube Shorts & Long-Form: CTR evaluation and pacing optimization
    • X (Twitter): High-bookmark utility threads and polarizing controversy hooks

    • Script & Strategy Deliverables Only: Generates structured multi-platform video scripts, scene visual directions, pacing timecodes, and audio cues, but does not render or export compiled MP4/MOV video files.
    • No Unauthorized Live Scrapers: Does not ship with fragile headless browser scrapers to prevent creator IP/account bans; trend ingestion relies on user inputs or the host agent's native browsing tools.
    • Spoken Pacing Calibration: Word-per-minute cadence algorithms (150–170 WPM) and punctuation density metrics are specifically calibrated for English spoken pacing across short-form platforms.
    • Single-Device Local Storage: The SQLite vault is designed for single-user local disk persistence (~/.consigliere/) rather than distributed multi-tenant cloud sync.

    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 any payment.

    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 1 month ago

    Requires

    Python 3.8+
    Python

    Permissions

    Terminal / Shell
    Network Access

    File Scopes

    ~/.consigliere/**
    scripts/**
    references/**
    templates/**
    **/*.md
    **/*.txt
    **/*.json

    Permissions are strictly scoped for local execution and zero-cloud privacy: 1. `~/.consigliere/**`: Required to initialize and query the private local SQLite database (`vault.db`) storing creator persona profiles, topic records, and memory logs on the user's machine without external telemetry. 2. `scripts/**`, `references/**`, `templates/**`: Read-only access to execute the packaged Python standard library algorithms, audit rubrics, and JSON schemas. 3. `**/*.{md,txt,json}`: Workspace read/write access to ingest raw creator transcripts/captions and export production-ready分镜 scripts, hook variations, and audit reports. No access to system files, network keys, or directories outside these boundaries.

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    The skill eliminates the "generic AI bot" tone through a 4-tier deterministic architecture rather than relying on standard LLM prompts: 1. Reverse-Engineered Empirical Cadence (Zero Generic Surveys): Instead of asking vague questions ("What is your tone?"), `profile_extractor.py` scans your actual past transcripts to extract mathematical speech metrics: median sentence length (e.g., 8.2 words vs. typical AI's 20+ words), period density, rhetorical question ratios, and specific opening rhythms. The generated draft is physically constrained to your speech cadence. 2. Hard-Coded AI Cliché Eliminator: Our `persona_auditor.py` applies strict regex filtering against known corporate AI clichés that instantly trigger viewer bounce: phrases like "In today's fast-paced world", "Delve into", "A testament to", "Game changer", "Dive right in", "Needless to say", and "In conclusion" are strictly flagged and eradicated. 3. "Electric Fence" Banned Words Enforcement: Your private SQLite profile maintains a customizable "Electric Fence" dictionary of words, jargon, and stylistic traits you personally despise. Any draft that breaches this fence is flagged with precise line-by-line violation reports. 4. Multi-Dimension Post-Generation Audit: Every script must pass an objective 4-dimension consistency audit (Cadence Adherence, Cliché Count, Banned Words, and Platform-Native CTA) before it reaches you, returning an empirical score (e.g., 96.5/100). If the score drops below 85, the engine flags exactly which sentences broke character.

    Yes, absolutely. The skill is specifically architected for cross-platform content adaptation across TikTok, YouTube (Shorts & Long-form), and X (Twitter): - Platform-Native Funnel Adaptation: Rather than blindly copy-pasting text, the engine recalculates the structural hooks based on distinct platform mechanics: • TikTok: Optimized for high Share/View ratios, dynamic visual pattern interrupts, and 15–45s rapid retention pacing. • YouTube Shorts & Long-Form: Structured around P25–P75 sweet-spot duration formulas, CTR hook packaging, and narrative retention curves. • X (Twitter): Tailored for high Bookmark-Rate contrarian opening hooks, concise scannable line breaks, and conversation-generating CTA closings. - Unified Persona Consistency: Ensures your core voice, cadence, and boundary rules remain 100% identical and instantly recognizable across all three networks.

    You receive a fully self-contained, production-grade package with zero external dependencies: 1. Core Directives: • `SKILL.md`: Standard AgentSkills/Agensi specification defining the 3-phase execution SOP. 2. Python Standard Library Tool Suite (`scripts/`): • `analytics.py`: 6 deterministic algorithms (1of10 Outlier Multiplier, Sweet Spot Duration, Platform Funnels, Trend Velocity). • `profile_extractor.py`: Reverse-extracts voice cadence, punctuation density, and sentence limits from raw transcripts. • `persona_auditor.py`: Post-generation 4-dimension audit (banned words, cadence length, AI cliché detection). • `vault.py`: Local private SQLite storage manager (`~/.consigliere/vault.db`) for profile and topic history. 3. Engineering References & Specs (`references/`): • `algorithms_spec.md`: Complete mathematical definitions, baseline tables, and edge-case handling. • `audit_rubric.md`: Scoring thresholds, deduction weights, and remediation protocols. • `hotspot_guide.md`: Trend ingestion, velocity phases, and blue ocean supply-demand matrix. • `persona_spec.md`: 6-slot persona parameter definitions and voice guardrail rules. 4. Schemas, Tests & Assets: • `templates/persona_schema.json`: Validated JSON schema for custom persona configurations. • `tests/test_preflight.py`: Preflight verification test suite (8/8 passing). • `assets/`: 512×512 high-resolution logo in both PNG and lossless SVG formats.

    The skill is engineered to be as friction-free and lightweight as possible: - Runtime: Python 3.8 or higher. - Dependencies: Zero external `pip` packages required. Runs 100% on native Python standard libraries (`sqlite3`, `math`, `re`, `json`, `pathlib`, `argparse`, `collections`). - Storage: Standard local disk write permission to create the private SQLite database at `~/.consigliere/`. - Operating System: 100% cross-platform compatible (Windows 10/11 with PowerShell/CMD, macOS Apple Silicon/Intel, and Linux). - Agent Harness: Works out-of-the-box with any standard agent runner or environment (Antigravity, Claude Desktop, Cursor, AgentSkills runners, or directly via CLI commands).

    No. Absolutely not. The skill requires zero access to your social media accounts, zero login credentials, zero OAuth tokens, and zero paid API keys. 1. 100% Local & Offline Computation: All 6 quantitative algorithms (1of10 Outlier Multiplier, Sweet Spot Duration, Engagement Funnels) and the Persona Consistency Auditor execute entirely on your machine using Python standard library. 2. Zero Cloud Telemetry: Your creator voice specs, past transcripts, and topic records are stored exclusively in an offline local SQLite database (`~/.consigliere/vault.db`). No data is ever transmitted to external cloud servers. 3. Zero Account Risk: You interact with the skill simply by feeding raw transcripts or topic ideas into your AI agent's chat or workspace. There is zero risk of platform strikes, API rate limits, or account bans.

    New in Agents & Orchestration