Add Prompt Injection Guard Skill

[DhanushNehru]

Create a skill that analyzes user inputs for prompt injection attempts and helps developers build safer AI applications. This is a unique, high-value security skill.

What it should do

• Analyze user inputs for common prompt injection patterns (ignore previous instructions, system prompt extraction, jailbreaks)
• Classify injection attempts by type: direct injection, indirect injection, prompt leaking, role hijacking
• Suggest input sanitization and guardrail strategies
• Generate test cases for red-teaming your own AI applications
• Provide a "safety score" for a given input

Why this is important

Prompt injection is the #1 vulnerability in AI applications (OWASP LLM Top 10). A skill that helps developers identify and defend against it is extremely valuable and shareable.

Skill File Location

skills/security/prompt-injection-guard.md

Acceptance Criteria

• Follows SKILL_TEMPLATE.md
• At least 3 example interactions covering different attack types
• Includes a classification taxonomy of injection types
• Tips section with defense-in-depth strategies

Difficulty: ⭐⭐⭐ (Advanced) — Requires AI security knowledge