The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read external context files, creating a surface for indirect prompt injection where malicious instructions could be embedded in the product marketing data.
Ingestion points: SKILL.md (Initial Assessment section) directs the agent to read .agents/product-marketing.md, .claude/product-marketing.md, or product-marketing-context.md.
Boundary markers: The instructions lack explicit delimiters or warnings to ignore embedded instructions within the ingested context files.
Capability inventory: The skill is instructional and does not define specific high-privilege tool invocations, but operates within an agent environment that typically possesses file and shell access.
Sanitization: There is no mention of sanitizing or validating the content of the ingested files before processing.

onboarding