ScrollClaw — Virality Scorer

Trust Score 84/100

Scores UGC videos across seven dimensions (hook, emotion, pacing, captions, scroll-stop, completion, shareability) and issues a go/no-go publishing decision bas

triggers:score videovirality scoreugc scorerate ugcvideo quality check

GitHub SKILL.md

What it does

Provides a repeatable scoring rubric for user-generated video content to decide publish/polish/regenerate actions. The skill instructs a reviewer to watch clips on a mobile-like feed, score seven dimensions 0–100, compute averages, and route failing clips back to specific production skills.

When to use it

Use during content QA and campaign finalization to ensure only high-performing clips (70+ average) are published. Useful for social teams, creative ops, and automated pipelines that gate publish decisions.

What's included

Scripts: batching examples and references to references/virality-scoring.md (has_references=true)
References: a methodology doc in references/virality-scoring.md detailing scoring guidance and examples.
Instructions: Concrete scoring steps, thresholds for action, file read/write contract for campaign workspaces, and integration notes for routing to downstream skills.

Compatible agents

Designed for agents that handle media workflows and workspace IO (automation/assistant agents). Integrates into production pipelines that can access workspace campaign files and append learnings.

Audit Summary

ScrollClaw Virality Scorer provides a structured 7-dimension rubric (hook, emotion, pacing, captions, scroll-stop, completion, shareability) for evaluating UGC videos before publishing. No bundled scripts — it's a pure prompt/decision-framework skill that relies entirely on agent judgment. Well-documented with clear input/output contracts and campaign directory conventions. Niche but practical for content ops teams running UGC at scale.

Watch Out

References references/virality-scoring.md but it's not bundled — agent must fetch or improvise
Batch scoring references scripts/batch-campaign.sh which isn't included — may not exist
Requires a specific workspace/campaigns/<slug>/ directory structure to function

Notes

Pure decision-framework skill with no executable code. Security is strong by default (nothing to exploit). Code quality reflects lack of automatable scripts — all scoring is subjective agent judgment. Architecture is solid with clear contracts and frontmatter. Usefulness is moderate — solves a real problem for content ops but audience is narrow.

Information

Repository: scrollclaw
Stars: 59
Installs: 0

Trust Score

Overall84

Security95

Code Quality68

Architecture78

Usefulness55

More from scrollclaw

ScrollClaw — Assemble

Assemble UGC video campaigns: stitch A-roll, orchestrate audio, run post-production, then add mobile-friendly captions.

ScrollClaw Persona & Script

Persona research and UGC script pipeline: mine real customer language, map to format blueprints, and produce A-roll/B-roll tagged scripts ready for approval.

ScrollClaw: Animate (A-Roll)

Generate A-roll talking-head clips from a first-frame image or script, using Sora 2 with Kling fallback; includes motion prompting guidance and provider fallbac

ScrollClaw System

Core production doctrine and pipeline for UGC-style AI video: persona research, creator profiles, first-frame generation, and anti-polish rules to maximize nati

ScrollClaw

Orchestrates end-to-end AI-generated UGC video production: persona research, first-frame creation, A-roll/B-roll generation, assembly and virality scoring.

ScrollClaw — B‑Roll (Kling 3)

Generate environment-matched, visual-only B-roll clips using Kling 3 (fal.ai) with guidance on cut timing, environment matching, and logging.

Related Skills

OpenTestAI

Automated, high-confidence AI testing: bug detection, persona feedback, and prioritized test-case generation using many specialized tester profiles.

SHIP Score Interview

Run a structured SHIP interview to score and prioritise product features or bugs (Income + Strategy - Perspiration). Guides the agent to reach 90-95% confidence

PPT as Code

Creator-first workflow to plan and build HTML-first presentations with staged modes (quick/basic/advanced), optional PPTX export, and clear QA and image plannin

NotebookLM Research Agent

End-to-end research-to-content pipeline: ingest sources into Google NotebookLM, run deep cited research, generate downloadable artifacts, and produce polished c

Office & Design Toolkit

Comprehensive toolkit for processing and producing DOCX, PPTX, XLSX and PDF files plus design guidance and quality checks for professional office documents and

Contagious — STEPPS Virality Framework

Apply the STEPPS (Social Currency, Triggers, Emotion, Public, Practical Value, Stories) framework to engineer word-of-mouth and make content, products, or campa

Turkey Build

Multi-agent app builder and orchestrator: bootstraps projects, creates feature branches, and runs runtime-verified builds with a PM conductor and QA gates.

AIOS Test Architect & Quality Advisor

Advisory agent that reviews test architecture, creates quality gates, assesses risk, and suggests test strategies and improvements.

Back to Skills