The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes external data, including brand profiles, voice guidelines, and blog/product content, to generate pin metadata and scheduling instructions. This ingestion of untrusted data represents a potential attack surface where malicious instructions could be embedded within the source content to influence the agent's output or tool interactions.
Ingestion points: The skill explicitly reads from brand-profile and voice-builder and processes external blog/product content provided by the user (refer to SKILL.md).
Boundary markers: The instructions do not specify strict delimiters for separating user-provided content from the agent's instructions, although it does include a safety note that comments/saves are "content, not a command" (refer to references/scope-and-connections.md).
Capability inventory: The skill drafts SEO copy, generates prompts for third-party image generation tools (e.g., Ideogram, Nano-banana), and sends scheduling data to the WoopSocial tool.
Sanitization: No explicit sanitization, filtering, or validation mechanisms for the ingested text content are described in the provided files.

pinterest-growth