The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to read from multiple external context files to inform its strategies, which creates a surface for indirect prompt injection if those files are compromised.
Ingestion points: Specifically instructs the agent to access .agents/marketing-learnings.md, .agents/marketing-strategy.md, .agents/product-marketing.md, .claude/product-marketing.md, and product-marketing-context.md for existing strategy context.
Boundary markers: The skill does not define specific delimiters or provide instructions to treat the content of these external files as untrusted or to ignore potential commands embedded within them.
Capability inventory: The agent is expected to read these files and apply their content to its reasoning and drafting process for marketing strategies.
Sanitization: No sanitization, escaping, or validation of the ingested content is specified in the skill's instructions.
[NO_CODE]: The skill consists entirely of instructional Markdown and evaluation JSON files. It does not contain or execute any Python, JavaScript, or shell scripts, significantly reducing the risk of malicious code execution.

community-marketing