The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest untrusted user data which could potentially contain malicious instructions.
Ingestion points: Untrusted user input enters the agent context through the $ARGUMENTS variable defined in SKILL.md to serve as the blog brief.
Boundary markers: The skill does not employ delimiters (like XML tags or triple quotes) or explicit instructions to the agent to ignore embedded commands within the $ARGUMENTS content.
Capability inventory: The skill is configured with Read and Write tool access, which allows the agent to interact with the file system based on instructions generated from the brief.
Sanitization: There is no evidence of input validation or content sanitization to prevent adversarial instructions from influencing the agent's behavior during the content generation workflow.

blog-writer