The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted text to perform stylistic rewriting, which creates a potential surface for indirect prompt injection. However, the agent's behavior is restricted by a comprehensive set of stylistic rules (e.g., removing specific vocabulary, forbidding em-dashes, and enforcing sentence structure) that focus the transformation on linguistic patterns rather than executing commands. * Ingestion points: User-supplied text provided to the agent for processing as defined in the process section (SKILL.md). * Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands within the input text. * Capability inventory: The skill uses natural language processing and text generation only. It does not utilize external tools, network capabilities, or file system access. * Sanitization: No validation, filtering, or sanitization of the input text content is implemented.

unslop