skills.rest
CES-Ltd
CES-LtdOFFICIAL
Explore all skills by this author →
WebsiteGitHub

skill-security-framing

Validates URLs and wraps external content as untrusted data for secure analysis workflows.

28|3|Updated May 10, 2026
Search Tags:#security#content analysis#content validation#URL validation#content wrapping#untrusted sources
One-click install
npx skills add https://github.com/CES-Ltd/Lumi --skill skill-security-framing-ces-ltd
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skill-security-framing
Source: https://github.com/CES-Ltd/Lumi/tree/main/resources/seed/octopus/skills/skill-security-framing
Command: npx skills add https://github.com/CES-Ltd/Lumi --skill skill-security-framing-ces-ltd

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

It addresses the risk of executing malicious or untrusted external content by defining standardized security patterns for validation and content wrapping.

Core Features & Use Cases

  • Establishes URL validation rules to reject dangerous protocols and hosts.
  • Provides platform-specific transformations, such as converting Twitter URLs to use safe API endpoints.
  • Implements a security frame template to wrap all fetched content, marking it as untrusted and instructing analysis subagents to treat it solely as data.
  • Guides safe analysis workflows to prevent prompt injection, code execution, or data exfiltration when processing external sources.
  • Ideal for any workflow that fetches, analyzes, or displays external web content or media in AI systems to mitigate security threats.

Quick Start

Wrap external fetched content in security context before analysis and ensure all URLs are validated to prevent security breaches.

Frequently Asked Questions about skill-security-framing

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

FAQPage Schema
How do I validate untrusted URLs before fetching external content?▼

URL validation for untrusted sources rejects dangerous protocols and hosts while applying platform-specific transforms like FxTwitter to ensure safe API endpoint extraction. This prevents malicious code execution during external content retrieval.

How does content wrapping protect AI agents from prompt injection?▼

Content wrapping prevents prompt injection by enforcing a security frame template that marks fetched external data as untrusted, instructing analysis subagents to treat the information strictly as raw data rather than executable commands.

What is the best way to secure external web content analysis workflows?▼

Securing external web content analysis requires standardized security patterns for content validation and wrapping, mitigating risks of code execution, prompt injection, and data exfiltration when processing untrusted sources.

When do I need to apply security framing to external content?▼

Security framing is required whenever workflows fetch, analyze, or display external web content in AI systems, ensuring fetched data is wrapped in a security context before analysis to prevent security breaches.

Can I use platform-specific transforms to safely extract Twitter URLs?▼

Platform-specific transforms enable safe Twitter URL extraction by converting standard links to use safe API endpoints like FxTwitter, allowing secure URL extraction without directly processing untrusted external content.

What are the limitations of processing untrusted external content without validation?▼

Processing untrusted external content without validation exposes AI systems to prompt injection, code execution, and data exfiltration because unvalidated data bypasses security protocols and may contain malicious instructions or dangerous hosts.

Related Skills

Explore More Content & Communication Skills→
nyldnnyldn
4.0k

skill-security-framing

Apply security framing to external content fetch and analysis workflows.

Community
Advanced
ManfronEnricoManfronEnrico

data-eda

Detect scientific file formats and generate Markdown analysis reports.

Community
Advanced
mhdxbilalmhdxbilal
1

skill-security-framing

Apply security framing to untrusted external content for safe analysis.

Community
Advanced
yhy0yhy0
74

vuln-patterns-deserialization

Identifies CVE-502 deserialization vulnerabilities in Python code, including untrusted sources and sinks like pickle.loads, yaml.load, torch.load, ZeroMQ, and json with hooks.

Community
Advanced