The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest external context from local files, which creates a potential surface for indirect prompt injection.
Ingestion points: In SKILL.md, the agent is directed to check for and read files such as .agents/product-marketing.md, .claude/product-marketing.md, or product-marketing-context.md before generating collateral.
Boundary markers: The instructions do not include specific delimiters or "ignore embedded instructions" warnings to mitigate the risk of the agent following malicious commands contained within those context files.
Capability inventory: The skill is limited to text generation for collateral and does not possess capabilities for command execution, file writing, or network operations.
Sanitization: There is no logic provided to sanitize or validate the content retrieved from the external context files before it is processed by the agent.
[NO_CODE]: The skill is composed exclusively of Markdown instructions and reference templates along with a JSON evaluation file. It does not include any Python scripts, Node.js packages, or other executable binaries, which significantly reduces the overall risk profile.

sales-enablement