The Agent Skills Directory

[SAFE]: The skill is a collection of instructional guidelines and templates designed for content strategy and copywriting. It does not contain executable code, scripts, or system-level commands.
[DATA_EXPOSURE]: The skill instructions include a step to read local context files such as .agents/product-marketing.md or .claude/product-marketing.md. These files are standard repositories for an agent's domain knowledge and do not contain sensitive system credentials or private user data.
[PROMPT_INJECTION]: The instructions do not contain patterns intended to override AI safety guardrails, hide system prompts, or bypass ethical constraints. The content is focused entirely on the stated purpose of competitor research and page generation.
[REMOTE_CODE_EXECUTION]: There are no indicators of external package installations, remote script fetches (e.g., curl|bash), or dynamic code execution patterns. All references are to internal documentation files provided within the skill package.
[INDIRECT_PROMPT_INJECTION]: While the skill ingests external data (competitive research from reviews and websites), the capability is limited to text generation for marketing copy. There are no exploitable capabilities such as file writing, shell access, or network requests that would allow for a malicious payload to be executed via indirect injection.

competitors