The Agent Skills Directory

[SAFE]: The skill is entirely instructional and provides strategic guidelines for marketing content. It does not contain any executable scripts, shell commands, or external dependencies that pose a security risk. The focus is on research-based content generation and SEO best practices.\n- [DATA_EXFILTRATION]: The skill instructions in SKILL.md direct the agent to check for and read .agents/product-marketing-context.md or .claude/product-marketing-context.md. While these files may contain sensitive product strategy, they are accessed within the local project environment for the purpose of personalizing generated content. No network-enabled tools or instructions for external transmission of this data were identified.\n- [PROMPT_INJECTION]: The skill establishes an ingestion surface for untrusted data by incorporating user-provided competitor information and local context files into generated marketing copy.\n
Ingestion points: Local marketing context files and user prompts specifying competitor details.\n
Boundary markers: The templates do not currently define explicit boundary markers to separate internal instructions from ingested data.\n
Capability inventory: The skill has no risky capabilities, such as subprocess execution, file system modification, or network access.\n
Sanitization: No specific input sanitization or output validation is described, which is typical for text-generation-only skills. The lack of sensitive capabilities renders this surface low-risk.

competitor-alternatives