The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read local project context files to tailor its recommendations. While this is intended functionality, content within these files could potentially influence agent behavior if the files are modified by untrusted sources.
Ingestion points: The agent is instructed to read .agents/product-marketing.md, .claude/product-marketing.md, or product-marketing-context.md in the skill root or subdirectories.
Boundary markers: The skill does not define specific delimiters or instructions to disregard embedded prompts within these ingested context files.
Capability inventory: The skill's capabilities are restricted to providing text-based marketing advice; it does not utilize shell execution, file-writing tools, or network access.
Sanitization: There is no logic provided to sanitize or validate the content retrieved from the context files before it is processed by the agent.

marketing-ideas