The Agent Skills Directory

[SAFE]: The skill's instructions are focused on content formatting and creative generation. No malicious patterns, obfuscation, or unauthorized system access commands were detected.- [DATA_EXPOSURE]: The skill reads from social-context.md and .agents/social-context.md to align with brand voice and audience requirements. This behavior is a standard functional requirement for the skill's purpose and does not involve accessing sensitive system credentials or protected directories.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from user source material and external context files.
Ingestion points: Content is read from social-context.md, .agents/social-context.md, and user-provided source material.
Boundary markers: None are explicitly defined in the prompt template.
Capability inventory: The agent is limited to text generation; no high-risk capabilities such as network operations, shell execution, or file writing are enabled or requested.
Sanitization: None detected. Despite the ingestion surface, the lack of dangerous tools or downstream execution capabilities results in a safe assessment.

carousel