The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest product marketing context from local files, which presents a potential surface for indirect prompt injection if those files contain untrusted content.
Ingestion points: Reads .agents/product-marketing.md or .claude/product-marketing.md (as identified in SKILL.md) to gather business context.
Boundary markers: The instructions do not define specific delimiters or warnings for the agent to ignore potentially malicious embedded instructions within these context files.
Capability inventory: The skill instructions focus on text generation for marketing strategies, sequence designs, and SMS copy; no high-privilege tool capabilities such as shell execution, file system modification, or network operations are requested or used by the skill.
Sanitization: There is no mention of sanitizing, escaping, or validating the content extracted from the context files.

sms