recursos-de-video-formula100k

• Outcome: Review
• Latest audit: Sep 17, 2026 · 4:19 PM UTC
• Version: 1.0.0

Overview

The skill mostly matches its video-resource purpose, but it should be reviewed because it may upload personal face photos to an external service and contains conflicting OpenClaw instructions about downloads and delivery.

Review this skill before installing if your videos, screenshots, or selfies are private. Use it only when you are comfortable with web searches, Telegram delivery, and Higgsfield processing, and require explicit confirmation before any personal face photo is uploaded. In OpenClaw, follow the header rules over the older generic steps: do not download TikToks with yt-dlp, do not use ffmpeg for TikTok/frame processing, and deliver the zip through Telegram.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (6)

Intent-Code Divergence

High

• Confidence: 98% confidence
• Finding: The OpenClaw-specific guidance explicitly forbids downloading TikToks with yt-dlp and extracting frames with ffmpeg, stating only links and cover images should be included. Later sections of the same file direct the agent to download MP4s via yt-dlp and process them with ffmpeg, which is a direct contradiction in documented intent versus described behavior.

Missing User Warnings

High

• Confidence: 97% confidence
• Finding: The workflow uploads user selfies/personal photos to Higgsfield as reference images without an explicit privacy disclosure or consent gate. Personal facial imagery is highly sensitive, and sending it to a third-party generation service can expose biometric or identity-related data beyond the user's local environment.

External Script Fetching

High

• Category: Supply Chain

Content

width: 1920
    height: 1080
    data:
      command: "curl -fsSL https://claude.ai/install.sh | bash"
      title: "Terminal — bash"
      slug: "curl-claude-install"
  - id: r02

• Confidence: 90% confidence
• Finding: Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The skill instructs packaging outputs and sending them over Telegram, an external messaging platform, without requiring explicit user consent or warning that video-derived content may leave the local workspace. That creates a real privacy and data-handling risk, especially if the materials contain personal, client, or unpublished information.

Intent-Code Divergence

Medium

• Confidence: 88% confidence
• Finding: The top-level OpenClaw instructions say this skill should work from an SRT or transcript and that Paso 2 does not run here. However, later sections still say activation requires a local video file and document a transcription/cut step producing edited assets, creating conflicting guidance about what inputs and processing this skill actually performs in this environment.

Intent-Code Divergence

Medium

• Confidence: 97% confidence
• Finding: The OpenClaw header specifies that delivery should be a zip sent over Telegram and explicitly says 'en vez de `open`'. Later, the delivery step still includes opening the destination folder locally, which contradicts the stated operational flow for this environment.

Static analysis

No suspicious patterns detected.