The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user data (career history, professional achievements, and project details) to generate branding documents. While this is necessary for its primary purpose, it creates an attack surface for indirect prompt injection if the user input contains malicious instructions.
Ingestion points: The '사전 수집: 개인 브랜드 인터뷰 체크리스트' section in SKILL.md requests detailed user information.
Boundary markers: No specific delimiters or 'ignore embedded instructions' warnings are used when processing user input.
Capability inventory: The skill produces five Markdown files (self_assessment.md, brand_strategy.md, brand_identity.md, content_strategy.md, channel_strategy.md) and chains to other skills for text refinement.
Sanitization: No explicit sanitization or validation of user-provided strings is mentioned.
[SAFE]: The skill uses internal tool references (e.g., moai-coworker, moai-writer) that align with the author's identity ('modu-ai'), representing standard modular functionality within the vendor's ecosystem.
[SAFE]: No obfuscation, unauthorized network operations, or persistence mechanisms were found across the instruction files or references.

marketing-personal-branding