The Agent Skills Directory

[SAFE]: The skill is composed of informational markdown files and templates. It does not contain executable scripts, shell commands, or external code dependencies.
[NO_CODE]: No functional code (Python, JavaScript, etc.) is included in the skill. All instructions are provided in natural language within markdown files.
[PROMPT_INJECTION]: The skill defines a surface for indirect prompt injection by instructing the agent to read .agents/product-marketing-context.md for context. This is a common pattern for context gathering and does not involve unsafe execution of the data. Evidence chain:
Ingestion points: .agents/product-marketing-context.md (referenced in SKILL.md)
Boundary markers: None explicitly defined in the instructions for reading the context file.
Capability inventory: The skill provides no additional scripts or tools; it relies on the agent's built-in file reading and search capabilities.
Sanitization: No validation or sanitization of the input file content is specified.

ai-seo