video-content-engine

Diagnose videos and produce validated multi-format content portfolios with delivery manifests.

3.5k|678|Updated Mar 28, 2026
One-click install
npx skills add https://github.com/ericosiu/ai-marketing-skills --skill video-content-engine
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: video-content-engine
Source: https://github.com/ericosiu/ai-marketing-skills/tree/main/video-content-engine
Command: npx skills add https://github.com/ericosiu/ai-marketing-skills --skill video-content-engine

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Turning one video into a coherent set of long-form, short-form, and written assets usually involves guesswork about what to cut, which formats to produce, and whether edits respect the source. This Skill replaces that guesswork with a source-grounded diagnosis, scored opportunity inventory, and validated delivery package.

Core Features & Use Cases

  • Source Diagnosis and Scoring: Scores hook, clarity, proof, pacing, and payoff from 0-20, then recommends the least destructive editorial mode (Polish, Tighten, Re-architect, or Rebuild).
  • Opportunity Routing: Builds a timestamped content-atom inventory and routes each atom to long-form, mid-form, short-form, micro, carousel, or written formats with Produce, Repair, Defer, or Reject verdicts.
  • Edit Validation: Ships Python validators that check clips begin and end on complete thoughts against word- or sentence-level transcripts, and verify delivery manifests with SHA-256 hashes.
  • Use Case: Given a 45-minute webinar recording, diagnose it, receive a recommended first release wave (e.g., one polished long-form cut, three Shorts, one carousel), produce only the approved modules, and validate the final delivery folder before publishing.

Quick Start

Use $video-content-engine to diagnose this authorized video at the provided URL and recommend the highest-leverage first release wave before producing anything.

Frequently Asked Questions about video-content-engine

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I turn one video into Shorts, Reels, and written content?

Provide an authorized video URL or file and the Skill builds a timestamped content-atom inventory, scores each atom, and routes complete atoms to short-form, micro, carousel, or written formats. It recommends a bounded first release wave rather than a fixed quota of derivatives.

How do I audit a video before editing it?

Run the Skill in its default diagnose mode, which scores hook, clarity, proof, pacing, and payoff from 0-20 and recommends an editorial mode: Polish, Tighten, Re-architect, or Rebuild. No production occurs until you approve specific modules.

What tools are required to run video editing workflows?

You need Python 3.10 or later, FFmpeg and FFprobe for media rendering and technical QC, and a transcription tool with word- or sentence-level timestamps. The bundled validation scripts use only the Python standard library.

Can the Skill publish or upload videos automatically?

No. Publishing, scheduling, uploading, changing sharing settings, spending quota, or activating campaigns all require explicit approval and an authoritative readback. The Skill defaults to diagnosis and never claims a render or publication without fresh evidence.

Why does edit boundary validation fail on my clips?

Validation fails when clip start or end times do not align with complete-thought boundaries in the exact-source transcript within the 0.12 second tolerance, or when the transcript lacks word or sentence granularity. Memo timestamps are treated as guides, never literal cut points.

What happens if a video link requires authentication or has DRM?

The Skill requests an accessible authorized source instead of analyzing from metadata alone. It never substitutes a mirror, alternate upload, or different episode, and it refuses to make editorial recommendations without grounded access to the media or transcript.