Mediaclaw Skill

MediaClaw is an AI video content factory for teams that need brand video production and repurposing at scale. It turns one source video into localized clips,...

Security Scan

The skill's claims and runtime instructions are high-level marketing for a video production pipeline and do not request unrelated credentials or install artifacts; overall the pieces are internally consistent though under-specified.

Purpose & Capability

The name, description, and SKILL.md all describe an AI video production pipeline (brand replacement, voiceover, subtitles, multi-platform exports). Nothing in the metadata (no env vars, no binaries, no install) contradicts that purpose. References to tools like demucs, Pillow, and TTS are plausible for this domain.

Instruction Scope

SKILL.md is high-level/marketing-oriented and does not include concrete runtime commands, file paths, or requests for local credentials. It asks for a source video and assets but does not tell the agent to read arbitrary system files or environment variables. Because instructions are non-prescriptive, the skill leaves implementation/IO behavior unspecified (where processing occurs, what endpoints are called).

Install Mechanism

There is no install specification and no code files to write to disk; this is instruction-only, which minimizes on-disk risk.

Credentials

The skill requests no environment variables or credentials. However, the described features (cloud TTS, face-swap, external model APIs) would normally require credentials or service endpoints in a real implementation; the absence of required creds simply means the skill is underspecified rather than requesting excessive access.

Persistence & Privilege

always is false and there are no install hooks or requests to modify agent/system config. The skill does not request persistent privileges.

Assessment

This skill appears coherent as a high-level description of a video-production assistant, but it is underspecified: it doesn't say where heavy processing happens or what external services it will call. Before installing or using it, verify: (1) who runs the processing (local vs. upload to mediaclawbot/mediaclaw.ai), (2) the privacy policy and data retention for uploaded videos, (3) what external API keys or services will actually be required, and (4) the vendor identity (there's no homepage in the registry metadata). Test with non-sensitive sample videos first and avoid providing private content or credentials until you have clear documentation of data flows and endpoints.

MediaClaw

MediaClaw is an AI-powered video content factory and media production platform that transforms one source video into dozens of branded, localized, platform-ready clips. Capabilities include brand video replacement, face swap, AI subtitle overlay, voiceover replacement, video repurposing, batch video creation, ad creative generation, and social media content automation. Supports TikTok, Instagram Reels, YouTube Shorts, Douyin, and Xiaohongshu. Built for brands, MCN agencies, e-commerce sellers, and marketing teams who need video content at scale.

What This Skill Does

MediaClaw is a full-stack AI video production pipeline. Give it a source video and brand assets, and it handles the entire workflow: analyzing scenes, replacing brand elements (logos, packaging, text), generating new captions with culturally-adapted copy, swapping voiceovers with natural TTS, and outputting platform-optimized clips for TikTok, Instagram Reels, YouTube Shorts, Xiaohongshu, and Douyin.

Core Capabilities

1. Brand Video Replacement — Swap product packaging, logos, and brand elements in existing videos using AI inpainting (Gemini frame editing + Kling/Seedance i2v generation)
2. Batch Video Production — Turn 1 hero video into 15+ unique variants with different hooks, captions, color grades, and aspect ratios
3. AI Subtitle Engine — Generate scroll-stopping captions with blue-word strategy, emoji hooks, and platform-specific formatting (Pillow frame-by-frame rendering)
4. Voiceover Replacement — Separate BGM from voice (demucs), generate new VO with matched pacing (TTS), remix seamlessly
5. Video Deduplication — 6-layer anti-duplicate system (prompt mutation, color shift, speed variation, crop jitter, subtitle style rotation, fingerprint perturbation)
6. Copywriting Engine — Platform-native ad copy generation with trending hooks, CTAs, comment-bait phrases, and de-duplication against posting history
7. Multi-Platform Export — Auto-resize and format for each platform's specs (9:16, 1:1, 16:9) with platform-specific watermark placement

Who It's For

• Brands managing social media video content at scale
• MCN agencies producing content for multiple KOLs
• E-commerce sellers needing product demo videos across platforms
• Marketing teams running video ad campaigns
• Content creators repurposing long-form into short clips

How It Works

You: Here's our product video and new brand assets
MediaClaw: Analyzing 12 scenes... Replacing brand elements... Generating 15 variants...
→ 15 platform-ready videos (TikTok/Reels/Shorts/XHS) with unique captions, hooks, and color grades
→ Matching ad copy for each platform
→ Performance tracking dashboard link

Usage

Describe your video production need:

• "Replace the beer brand in this commercial with our product"
• "Generate 10 TikTok variants from this hero video with different hooks"
• "Create Xiaohongshu video ads for our skincare line"
• "Batch produce 20 short videos for our 618 campaign"

Upgrade

For production pipeline access, API integration, and enterprise batch processing, visit https://mediaclawbot.com

---

Built by the MediaClaw team — AI video production at scale